2016 Long-Term Reliability Assessment

Transcription

1 2016 Long-Term Reliability Assessment December 2016 NERC Report Title Report Date I

2 Table of Contents Preface... v Introduction... vi Executive Summary... vii Chapter 1: Reliability Issues... 1 Resource Adequacy...2 MISO...6 NPCC-New England...8 Texas RE-ERCOT...9 WECC-CAMX Single Fuel Dependency Nuclear Uncertainty Chapter 2: Probabilistic Analysis Sensitivity Model VER Capacity Contributions Chapter 3: Essential Reliability Services Frequency Support ERCOT Eastern Interconnection Primary Frequency Response Analysis Net Load Ramping ERCOT s Sample results Distributed Energy Resources Diminishing On-Peak Impact of DER Distributed Energy Resources Task Force Chapter 4: Reliability Assessment Trends and Analysis Reserve Margins Demand and Energy Demand-Side Management Generation Fuel Mix Transmission Adequacy FRCC MISO Manitoba Hydro ii

3 Table of Contents SaskPower Maritimes New England New York Ontario Hydro Québec SERC SPP Texas RE-ERCOT Chapter 5: Additional Reliability Issues Clean Power Plan System Short-Circuit Strength Short-Circuit Fault Contribution of Nonsynchronous Resources Nonsynchronous Plant Types System Strength Modeling Case Quality Metrics Dynamic Load Modeling Distributed Energy Resource Modeling Reactive Power Requirements for Nonsynchronous Generation: FERC Order System Restoration Reactive Power Supporting Devices Advanced Capabilities of New Technology Resources Inertia-based Controls Governor Control Solar Eclipse Chapter 6: Regional Overview Highlights of Emerging Issues Probability-Based Resource Adequacy Assessment Assessment Area Granular Review FRCC MISO MRO-Manitoba Hydro MRO-SaskPower NPCC-Maritimes iii

4 Table of Contents NPCC-New England NPCC-New York NPCC-Ontario NPCC- Québec Demand, Resources, and Planning Reserve Margins PJM SERC SPP Texas RE-ERCOT WECC Appendix I: Reliability Assessment Glossary Appendix II: Assessment Preparation, Design, and Data Concepts iv

5 Preface The North American Electric Reliability Corporation (NERC) is a not-for-profit international regulatory authority whose mission is to assure the reliability of the bulk power system (BPS) in North America. NERC develops and enforces Reliability Standards; annually assesses seasonal and long term reliability; monitors the BPS through system awareness; and educates, trains, and certifies industry personnel. NERC s area of responsibility spans the continental United States, Canada, and the northern portion of Baja California, Mexico. NERC is the electric reliability organization (ERO) for North America, subject to oversight by the Federal Energy Regulatory Commission (FERC) and governmental authorities in Canada. NERC s jurisdiction includes users, owners, and operators of the BPS, which serves more than 334 million people. NERC Regions and Assessment Areas FRCC Florida Reliability Coordinating Council FRCC MRO Midwest Reliability Organization MRO-SaskPower MRO-Manitoba Hydro MISO NPCC Northeast Power Coordinating Council NPCC-New England NPCC-Maritimes NPCC-New York NPCC-Ontario NPCC-Québec RF ReliabilityFirst PJM SERC SERC Reliability Corporation SERC-East SERC-North SERC-Southeast SPP RE Southwest Power Pool Regional Entity SPP Texas RE Texas Reliability Entity Texas RE-ERCOT WECC Western Electricity Coordinating Council WECC-BC WECC-AB WECC-RMRG WECC-CA/MX WECC-SRSG WECC-NWPP-US v

6 Introduction NERC prepares seasonal and long-term assessments to examine current and future adequacy and operational reliability of the North American BPS. For these assessments, the BPS is divided into 21 assessment areas 1 both within and across the eight Regional Entity boundaries as shown in the corresponding table and maps in the preface. 2 The preparation of these assessments involves NERC s collection and consolidation of data from the Regional Entities. Reference Case data includes projected on-peak demand and energy, demand response (DR), resource capacity, and transmission projects. Data and information from each NERC Region are also collected and used to identify notable trends, emerging issues, and potential concerns (see Chapter 6). This bottom-up approach captures virtually all electricity supplied in the United States, Canada, and the portion of Baja California Norte, Mexico. NERC s reliability assessments are developed to inform industry, policy makers, and regulators and to aid NERC in achieving its mission to ensure the reliability of the North American BPS. NERC s primary objective with the LTRA is to assess resource and transmission adequacy across the NERC footprint, and to assess emerging issues that have an impact on BPS reliability over the next ten years. NERC assesses this reliability by comparing projected reserve margins to Reference Margin Levels established by the assessment area or to a default Reference Margin Level. Reserve margins are typically developed using probabilistic methods that calculate the loss of load expectation (LOLE) that could occur less than or equal to one time in ten years based on daily peak information. Whereas these analyses typically evaluate resource adequacy in order to meet a peak day requirement. NERC recognizes that a changing resource mix with a significant portion of it being energy-limited, changes in off-peak demand, single points of disruption, and other factors can have an effect on resource adequacy. As a result, NERC is incorporating more probabilistic approaches into this assessment and other ongoing analyses that provide further insights into how to best establish adequate reserve margins amidst a BPS undergoing unprecedented changes. Additional issues that may potentially impact the reliability of the BPS, such as physical and cybersecurity, are not specifically reviewed by this assessment. These issues present constant and evolving challenges. NERC continues to lead a multi-faceted approach to enhancing cybersecurity, through mandatory standards, improved information-sharing through the Electricity-Information Sharing and Analysis Center (E-ISAC), 3 and exercises to increase learning about threats and vulnerabilities. NERC has prepared the following assessment in accordance with the Energy Policy Act of 2005 in which the United States Congress directed NERC to conduct periodic assessments of the reliability and adequacy of the BPS in North America. 4 This assessment is based on data and information collected by NERC from the Regions on an assessment area basis as of September The Reliability Assessment Subcommittee (RAS), at the direction of the Planning Committee (PC), supports the LTRA s development. Specifically, NERC and the RAS perform a thorough peer review that leverages the knowledge and experience of industry subject matter experts while providing a balance to ensure the validity of data and information provided by the Regions. Each assessment area section is peer reviewed by members from other Regions to achieve a comprehensive review that is verified by the RAS in open meetings. The review process ensures the accuracy and completeness of the data and information provided by each Region. This assessment has been reviewed and accepted by the PC. The NERC Board of Trustees also reviewed and approved this report. 1 The number of assessment areas has remained the same since the release of the 2015 LTRA. The previous MRO-MAPP footprint now resides in the SPP, MISO, and WECC-NWPP-US assessment areas. The previous WECC-CA has split into the WECC-BC and WECC-Alberta. 2 Maps created using ABB Velocity Suite. 3 NERC Electricity ISAC 4 H.R. 6 as approved by of the One Hundred Ninth Congress of the United States, the Energy Policy Act of The NERC Rules of Procedure, Section 800, further detail the objectives, scope, data and information requirements, and Reliability Assessment Process requiring annual seasonal and long-term reliability assessments. vi

7 Executive Summary The 2016 Long-Term Reliability Assessment (2016 LTRA) provides a wide-area perspective of the generation, demand-side resources, and transmission system adequacy needed during the next decade. This assessment includes NERC s independent technical analysis to identify issues that may impact North American bulk power system (BPS) reliability, and this enables industry, regulators, and policy makers to develop mitigation plans or strategies to address them. NERC collected projections from system planners in each assessment area, assessed the data independently, and then identified emerging issues. These reliability issues require consideration to reduce BPS reliability risks and are summarized here in six focus areas: Resource Adequacy: Factors that are included when performing a resource adequacy assessment include a reserve margin analysis and the study of emerging reliability issues that can impact generation and demand projections. The results of this study identified four assessment areas as having a medium resource adequacy risk in the first five years of the assessment period: MISO: The Anticipated Reserve Margin falls to 13.8 percent, below MISO s Reference Margin Level of 15.2 percent, in Due to the uncertain outcome of pending regulatory requirements in MISO s footprint, 3.3 GW of capacity are categorized as unconfirmed retirements. When applying these additional potential retirements, MISO s Anticipated Reserve Margin decreases to 14.9 percent in 2018, which is below the Reference Margin Level of 15.2 percent. NPCC-New England: Reserve margins in NPCC-New England are projected above the Reference margin level for all years of the assessment. However, an increased reliance on natural gas, coupled with limited gas storage capability and dual-fuel switching challenges, indicate a medium resource adequacy risk, particularly during the winter peak season. Texas RE-ERCOT: Anticipated Reserve Margins are projected to be sufficient for all ten years of the assessment period. Due to the uncertain outcome of pending regulatory requirements, approximately 7 GW of capacity are categorized as unconfirmed retirements. With considerations for unconfirmed retirements and assuming no potential replacement capacity, Texas RE-ERCOT s Anticipated Reserve Margin decreases to 11.3 percent by 2021, which is below the Reference Margin Level of percent. WECC-CAMX: Anticipated Reserve Margins are projected to be sufficient for all ten years of the assessment period. However, an increased reliance on natural gas, limited dual-fuel capability, and natural gas storage facility outages indicate a medium resource adequacy risk. Additional challenges are posed in maintaining adequate essential reliability services (ERSs), such as maintaining ramping capability. Single-Fuel Dependency: NERC has identified that reliance on a single fuel increases vulnerabilities, particularly during extreme weather conditions. Over the past decade, several areas have significantly increased their dependence on natural gas. This trend has continued amidst historically low natural gas prices and regulatory rulings that continue to promote increased natural gas generation. NERC s assessment identifies four assessment areas with high penetration of natural gas generation: Texas RE-ERCOT: Natural-gas-fired generation comprises 63 percent of on-peak anticipated capacity by Gas-fired generators in ERCOT have some dual-fuel capability (14 percent). However, pipeline infrastructure in Texas is tightly meshed, and natural gas generators often have multiple connections and access to natural gas. FRCC: Natural-gas-fired generation will comprise 69 percent of on-peak anticipated capacity by Natural gas is not widely used for residential heating; therefore, the pipeline system has largely been built to support its gas-fired generation customers. Gas-fired generation in Florida is largely fueled with firm transportation services that is approved by the public utility commission to ensure a reliable vii

8 Executive Summary source of fuel. Additionally, a majority of gas-fired generation is dual-fuel capable, and limited inventory is kept on-site for use in emergencies. NPCC-New England: Natural-gas-fired generation will comprise 52 percent of on-peak anticipated capacity by The risk in New England is increasing due to the limited addition of new interstate pipeline capacity and the fact that natural gas storage does not appear to keep pace with natural gas generation additions. Additionally, recent winter experiences have created challenges in both maintaining back-up fuel inventories and successfully switching from gas to oil. However, emerging market rules in ISO-NE, beginning in 2018, are expected to support reliability and the resilience of the generation fleet. WECC-CAMX: Natural-gas-fired generation comprises 68 percent of on-peak anticipated capacity by Minimal dual-fuel capable units and immediate resource constraints from the outage at the Aliso Canyon underground natural gas storage facility increase the risks associated with single-fuel dependency. Nuclear Uncertainty: Low natural gas prices continue to affect the competitiveness of nuclear generation and are a key contributing factor to nuclear generation s difficulty in remaining economic with competing fuel sources. While new nuclear facilities are being built in Georgia, Tennessee, and South Carolina, potential retirements have been announced for nuclear facilities in Illinois, California, Nebraska, Massachusetts, and New York, creating longer-term uncertainty for system operators and planners. While replacement capacity may be advanced to mitigate resource adequacy concerns, unconfirmed nuclear retirements create uncertainty around local transmission adequacy and the ability to plan for future resource and demand needs due to their large baseload contribution. Probabilistic Analysis: The changing resource mix introduces additional complexities to assessing resource adequacy and diminishes the value of a single deterministic planning metric (e.g., reserve margins). NERC s probabilistic assessment (see Chapter 2) provides key indices that, together, assess resource adequacy risks for all hours of the study year. Essential Reliability Services: The addition of a large number of variable energy resources (VERs) onto the BPS has resulted in the need for operational flexibility to accommodate demand while also effectively managing the resource portfolio. As VERs are becoming more significant, NERC is developing sufficiency guidelines in order to establish requisite levels of ERSs. ERSs are comprised of primary frequency response (PFR), voltage support, and ramping capability, all needed for the continued reliable operation of the BPS. Significant ramping capabilities are needed to address the challenges presented from VER operational impacts. Ramping issues requiring increased operational flexibility have been most notable in California, where they occurred four years earlier than originally projected. Texas RE-ERCOT is also beginning to project potential ramping issues, but current real-time market design, operating practices, and the flexibility of existing generation provide ERCOT with sufficient capability to manage ramping requirements. Distributed Energy Resources: Increasing installations of distributed energy resources (DERs) modify how distribution and transmission systems interact with each other. Many utilities currently lack sufficient visibility and operational control of these resources, increasing the risk to BPS reliability. This visibility is a crucial aspect of power system planning, forecasting, and modeling that requires adequate data and information exchanges across the transmission and distribution interface. The most significant growth in DER penetration is occurring in NPCC and WECC. NERC s Distributed Energy Resources Task Force will release their initial their report in Q1 of This report will review current impact to reliability and considerations for resource and transmission planning. viii

9 Executive Summary Recommendations NERC has developed the following recommendations through its stakeholder process to alleviate the potential impacts of the reliability issues identified in this assessment: Regulators and legislators should evaluate the changes occurring on the BPS irrespective of the final rulings on pending regulation, such as the Clean Power Plan (CPP). While there is uncertainty around the ultimate validity and timing of the CPP, NERC has determined that many of the changes are occurring regardless of the final ruling. As the resource mix continues to change, the need for more investments in transmission and natural gas infrastructure is currently projected. The lengthy schedule involved in acquiring, siting, and permitting adequate properties for this infrastructure should also be considered when assessing reliability impacts. Policy makers should closely monitor and evaluate the measures being taken to address the evolving resource adequacy trends in MISO and Texas RE-ERCOT. As natural-gas-fired resources continue to increase, system planners and operators should evaluate the potential effects of an increased reliance on natural gas on BPS reliability. Natural gas provides just-in time fuel; therefore, firm transportation and maintaining dual-fuel capability can significantly reduce the risk of common-mode failure and wider-spread reliability challenges. As part of future transmission and resource planning studies, planning entities will need to more fully understand how impacts to the natural gas transportation system can affect electric reliability. Regulators and legislators should consider the uncertainties in resource retirements and resource mix changes projected by resource planners and the interconnection-wide impacts, including generation retirements, curtailments, and transmission constraints that can manifest if ERSs are not maintained. The implementation of a regulatory framework to provide an adequate level of ERSs could help to address these uncertainties. Planning Coordinators and Transmission Planners should consider supplementing planning processes with additional measures that support maintaining sufficient ERSs. In 2017, NERC will draft sufficiency guidelines for ERSs to support planning evaluations and assessments of how the resource mix can impact BPS reliability; NERC recommends incorporating sufficiency measures within planning processes. ix

10 Chapter 1: Reliability Issues This section highlights several issues that are emerging and have the potential to increase risks to reliability. The 2016 LTRA identifies these issues to include: resource adequacy, single-fuel dependency, nuclear uncertainty, essential reliability services (ERSs), and distributed energy resources (DERs). Reserve Margins General Assumptions: The reserve margin calculation is an important industry planning metric used to examine future resource adequacy. This deterministic approach examines the forecast peak demand (load) and projected availability of resources to serve the forecast peak demand for the summer and winter of the 10-year outlook ( ). Demand Assumptions: Electricity demand projections, or load forecasts, are provided by each assessment area. Load forecasts include peak hourly load, or total internal demand, for the summer and winter of each year. Total internal demand projections are based on normal weather (50/50 distribution) and are provided on a coincident basis for most assessment areas. Net internal demand equals the total internal demand minus the controllable & dispatchable demand response (DR) considered available across the peak. Resource Assumptions: NERC collects projections for the amount of existing and planned capacity as well as net capacity transfers (between assessment areas) that will be available during the forecast hour of peak demand for the summer and winter seasons of each year. Resource planning methods vary throughout the North American BPS. NERC uses the following categories to provide a consistent approach for collecting and presenting resource adequacy: Anticipated Resources Existing-Certain generating capacity: includes operable capacity expected to be available to serve load during the peak hour with firm transmission. Tier 1 capacity additions: includes capacity that has completed construction, is under construction, has a signed or approved ISA/PPA/CSA/WMPA, is included in an integrated resource plan, or is under a regulatory environment that mandates a resource adequacy requirement. Firm Capacity Transfers (Imports minus Exports): transfers with firm contracts Prospective Resources: Includes all anticipated resources, plus: Existing-Other capacity: includes operable capacity that could be available to serve load during the peak hour, but lacks firm transmission and could be unavailable for a number of reasons. Tier 2 capacity additions: includes capacity that has been requested but that has not received approval for planning requirements. Expected (non-firm) Capacity Transfers (Imports minus Exports): transfers without firm contracts, but a high probability of future implementation. Reserve Margins: the primary metric used to measure resource adequacy. It is defined as the difference in resources (anticipated, or prospective) and net internal demand divided by net internal demand, shown as a percent. (Anticipated Resources Net Internal Demand) Anticipated Reserve Margin = Net Internal Demand (Prospective Resources Net Internal Demand) Prospective Reserve Margin = Net Internal Demand Reference Margin Level: The assumptions of this metric vary by assessment area. Generally, the Reference Margin Level is based on load, generation, and transmission characteristics for each assessment area and, in some cases, the Reference Margin Level is a requirement implemented by the respective state(s), provincial authorities, ISOs/RTOs, or other regulatory bodies. If such a requirement exists, the respective assessment area generally adopts this requirement as the Reference Margin Level. In some cases, the Reference Margin Level will fluctuate over the duration of the assessment period, or may be different for the summer and winter seasons. If one is not provided by a given assessment area, NERC applies a 15% Reference Margin Level for predominately thermal systems and 10% for predominately hydro systems. 1

11 Chapter 1: Reliability Issues Resource Adequacy The Anticipated Reserve Margin is the primary metric that is used to evaluate the adequacy of projected resources to serve forecasted peak load. Figure 1.1 provides an examination of the Anticipated Reserve Margin and how to interpret the results of the analysis. Having a shortfall of reserves indicates that an assessment area would fall below their target Reference Margin Level, and increases the risk to reliability by increasing the likelihood of a potential loss of load. Based on the data and information provided to NERC, all assessment areas Anticipated Reserve Margins that meet or exceed their Reference Margin Levels. While three areas fall below their respective Reference Margin Levels in the 6- to 10-year time frame, there are measures that can be taken to address potential shortfalls. Examples include advancing designated planned resources within the generation queue, securing neighboring capacity through Figure 1.1: Examination of Anticipated Reserve Margin transmission expansions, and firm transmission contracts. Generally, shortfalls identified in the latter years of the assessment period pose a less significant risk to resource adequacy due to more time and available mitigation options. Alternatively, an assessment area with additional planning reserves may not maintain the requisite level of ERSs, thereby introducing complexity into an assessment of resources to consider other measures of sufficiency in addition to reserve margins. 5 Figure 1.2 and Figure 1.3 show the five- and ten-year planning reserve margins respectively. All assessment areas meet or exceed their Anticipated Reserve Margins through the first five years of the assessment while three assessment areas Anticipated Reserve Margins do fall under their designated Reference Margin Level by year ten. Figure 1.2: Planning Reserve Margins for Year 5 (2021) 5 Requisite levels of ERSs will be determined by the Sufficiency Guidelines currently under development with the NERC Essential Reliability Services Working Group 2

12 Chapter 1: Reliability Issues MISO is currently projected to fall below their target of percent to an Anticipated Reserve Margin of percent in 2022 and continue to decrease to 9.07 percent by the year NPCC-Québec is currently projected to fall below their target of percent to percent in 2025 followed by a decrease to percent by WECC-BC is currently projected to fall below their target of percent to percent in 2025 followed by a decrease to 9.79 percent by Figure 1.3: Planning Reserve Margins for Year 10 (2026) Figure 1.4 below contains a qualitative risk process flowchart for the short-term (1-to-5 years) outlook and a chart showing the corresponding resource adequacy risk of a capacity shortage by assessment area. The 6-to-10 year outlook flowchart is a direct extension of the 1-to-5 year. The flowchart outcomes for this qualitative analysis result in flagging an assessment area as: 1) a high concern (shown by a red circle indicator), 2) a medium concern (shown by a yellow circle indicator), or 3) a low concern (shown by a green circle indicator). For example: an area is flagged as high concern when a Region s Anticipated Reserve Margin (ARM) is less than the Reference Margin Level (RML) in both the 1-to-5 year outlook and the 1-to-3 year time frame. When one or a combination of factors contribute to risk, the area is flagged as a medium concern. The factors that are considered are as follows: Prospective Reserve Margin (PRM), ARM, RML, unconfirmed retirements, and emerging and sustaining issues. Lastly, if an area is not flagged as high or medium, it is identified as having a low concern with respect to near- and long-term resource adequacy risk. Confirmed and Unconfirmed Retirements NERC collects two separate line items for retirements in the development of the Long-Term Reliability Assessment: Units whose retirements are designated as confirmed have formally announced plans to retire and these units must have an approved generator deactivation request where applicable. These units are individually identified within the data collection. Unconfirmed retirements are collected but aggregated by fuel type. These include units that have been earmarked for retirement but have not met the same requirements as those given as confirmed. These include units that are expected to retire based on the result of a generator survey or assessment area resource adequacy study. 3

13 Chapter 1: Reliability Issues Figure 1.4: Qualitative Risk Process Flowchart (left) and Corresponding Perceived Resource Adequacy Risk Results of this qualitative risk analysis indicate that a total of four assessment areas have a medium reliability concern in the short-term and six assessment areas for the long-term. Assessment areas that indicated a medium risk in the short term are reviewed in more detail. These include: MISO: Lower Anticipated Reserve Margins and a significant amount of unconfirmed retirements due to the potential outcome of pending regulatory requirements indicate a medium resource adequacy risk. NPCC-New England: A growing reliance on natural gas, the lack of dual-fuel compatible units, and limited gas storage capability indicate a medium resource adequacy risk. Texas RE-ERCOT: Significant amounts of unconfirmed retirements due to the potential impacts of pending environmental regulations and a high reliance on natural-gas-fired generation indicate a medium resource adequacy risk. WECC-CAMX: A high reliance on natural gas, limited dual-fuel capability, and the potential reduction in adequate ERSs indicate a medium resource adequacy risk. 4

14 Chapter 1: Reliability Issues Recent environmental and other regulatory requirements have introduced greater uncertainty around the future of some resources. In addition to the other challenges brought forward by the incorporation of a changing resource mix, this increasing uncertainty of future resources is compounded by advanced retirements of conventional fossil-fired generating units and the capacity contributions expected from an increasing amount of variable generation. Figure 1.5 shows the total amount of capacity retirements projected to occur in 2015 and 2016 using forecasted and actual data. The actual 2015 retirements totaled 24.3 GW, which was 3.3 GW more, or 15 percent, than the 21 GW forecasted by the 2015 LTRA. Similarly, the 2015 LTRA forecasted 9.3 GW of retirements to occur in 2016, but the 2016 LTRA estimates an additional 3.4 GW, or 36 percent, more than the prior year s assessment. Figure 1.5: Total Projected Confirmed Retirements Increase between Assessment Years As the outcome of retirement decisions are made public, upward adjustments on expected retirements continue to be incorporated into NERC s projections. Since 2012, approximately 40 GW of coal-fired and 30 GW of oil-fired generation have been retired in North America, roughly 7 percent of 2016 summer peak demand. NERC examines both confirmed and unconfirmed generation retirements in the 10-year forecast. The Anticipated Reserve Margin includes only generation retirements that have been confirmed. However, given federal, state, and provincial policies, a number of power plants have an increased risk of retirement, and this should be considered when evaluating planning reserve margins. NERC s assessment found both the MISO and Texas RE- ERCOT assessment areas showing significant changes to their reserve margin forecasts when considering unconfirmed retirements and assuming no potential replacement capacity. Figure 1.6 shows the total amount of actual nameplate capacity retirements from 2015 and the projected retirements for years by fuel type. The data indicates there were significant coal retirements in 2015, greater than the total retirements projected for the ten-year assessment period. 5

15 Chapter 1: Reliability Issues Figure 1.6: NERC-Wide MW Nameplate Capacity Retirements from 2015 to 2026 by Fuel Type *Actual Data 6 NERC also conducted a sensitivity analysis in which no Tier 2 capacity was built and all unconfirmed retirements were taken out of service. The aggregated unconfirmed retirements were provided from MISO through the Organization of MISO States (OMS) survey results for This provides insight on the potential retirement of many resources in the MISO footprint. The survey results provide a greater confidence factor to apply the unconfirmed retirements into a reserve margin sensitivity analysis. Similarly, ERCOT released their 2016 CDR, providing additional detail on power plant retirement risks and generation fleet changes. 8 While both MISO and ERCOT have sufficient Tier 2 resources in the queue, depending on the timing of the retirements (Tier 2 resources may not be available to advance their in-service dates), which could increase the risk of an electricity supply shortage. MISO Similar to the 2014 LTRA and 2015 LTRA reference cases, the 2016 LTRA reference case projects a shortfall in MISO s Anticipated Reserve Margins during the assessment period. The shortfall in projections is due to generation retirements outpacing the addition of Tier 1 resources; there is sufficient Tier 2 and Tier 3 generation that could be advanced to mitigate these capacity concerns. MISO is projecting an Anticipated Reserve Margin of 13.8 percent for the 2022 summer peak, which continues to trend downward to 9.0 percent by the end of MISO will require approximately 8 GW of additional resources by the end of the 10-year forecast in order to maintain the Reference Margin requirements of 15.2 percent. Considerations should be given to the assessment area's need for sufficient ERSs. These may include generation additions that are mostly asynchronous and may offer a reduced level of voltage, frequency, and/or ramping support, depending on equipment characteristics and facility design. Shown in Figure 1.7, the Reference Margin requirements are up by 0.9 percent compared to the 2015 LTRA reference case due to resource adequacy study assumptions. These changes are mostly due to the 6 Actual data for 2015 collected from EIA Electric Power Monthly 7 Organization of MISO States Survey Results; Report on the Capacity, Demand and Reserves (CDR) in the ERCOT Region, ; May

16 Chapter 1: Reliability Issues planning year being the first year of integrating the MISO South Zone with limited data being available. 9 Figure 1.7: MISO 2015 LTRA and 2016 LTRA Reserve Margin Comparison MISO gathered data for the past three years through the OMS Survey as part of their resource adequacy study. Survey results indicate that certain locations within the assessment area will have to rely on imports as early as 2017 from their neighboring zones, such as Missouri and Lower Michigan. The survey resulted in an estimation of 3.3 GW plant retirements by NERC considers these retirements as unconfirmed and are the major contributor in the advanced Reserve Margin shortfalls. ReliabilityFirst s 2016 Long-Term Resource Report also identified these potential risks highlighted by the OMS survey results. 10 Deliverability of New Resources One of the major challenges in long-term system planning is the changing nature and location of available resources to load. The North American BPS does not provide infinite routes for all generation; therefore, the transition from a central-station model to a more dispersed BPS creates some challenges in power delivery and transmission. System planners use modeling software to simulate current and projected grid components and characteristics. From these models, transmission planners will identify potential future contingencies on lines and evaluate options, such as uprating or building new lines to mitigate contingencies before they occur. Having new resources built long distances from the load requires that new lines be built to effectively deliver this new generation to where it is needed. Transmission congested lines and operational challenges are likely to escalate within an area if the constraints are not alleviated. Figure 1.8 includes the resulting unconfirmed retirement sensitivity analysis impacts on MISO s Anticipated Reserve Margins, which will fall below the Reference Margin Level by While the reference margin is not met in the five-year period given unconfirmed retirements, MISO appears to have sufficient Tier 2 resources to meet the Reference Margin Level. The long-term resource adequacy forecast is generally low risk, but as variable resources increase, Reference Margin Level requirements may increase beyond the current 15.2 percent in the future years. 9 MISO Loss of Load Expectation Study Report: Planning year ReliabilityFirst 2016 Assessment-Long Term Resource; August

17 Chapter 1: Reliability Issues Figure 1.8: MISO Reserve Margins with Unconfirmed Retirements MISO s long-term resource challenges are exacerbated by increasing transmission requirements. The MISO forecast includes a significant expansion of wind resources. Because of the geographic diversity of wind resources to load, more long-distance and networked transmission will be needed. Ensuring the deliverability of these resources is challenging when resources are located distant from the load. For example, forced curtailments of wind resources are sometimes required to prevent congestion on transmission lines. An August 2016 report by the U.S. Department of Energy 11 showed that the percentage of wind curtailment in MW to the total potential wind generation has increased in MISO from under 2 percent in 2007 to over 5.5 percent in An increase in wind curtailments could be a result of transmission inadequacy, minimum generation limits, other forms of grid inflexibility, and/or environmental restrictions. This could lead to an increased risk of real-time capacity deficiencies. 12 NPCC-New England The Anticipated Reserve Margins for NPCC-New England, shown below in Figure 1.9, exceed the Reference Margin Level for all years of the assessment period. Compared to the 2015 LTRA reserve margin analysis, the Anticipated Reserve Margins have increased by 0.5 percent in 2017 and by 3.32 percent by year The majority of this change is due to a slight reduction in the ten-year peak load forecast. Figure 1.9: NPCC-New England 2015 LTRA and 2016 LTRA Reserve Margin Comparison 11 Department of Energy: Wind Technologies Market Report - August Ibid. 8

18 Chapter 1: Reliability Issues The results from the qualitative risk analysis, as presented in the Qualitative Risk Process flowchart (Figure 1.4), highlighted NPCC-New England as having a perceived medium resource adequacy risk. While the results of the reserve margin analysis indicate that NPCC-New England has sufficient capacity to remain above their Reference Margin Level, there are other standing or emerging issues that must be considered when performing a more holistic overview of the assessment area s potential risks to reliability. NPCC-New England faces additional challenges due to a high dependency on natural gas, a reduced dual-fuel capable fleet, and limited storage capability. These challenges are exacerbated by high winter gas demand competitiveness from customers other than electric generating facilities and inadequate natural gas pipeline infrastructure. 13 Texas RE-ERCOT NERC s 2015 LTRA reference case showed Texas RE-ERCOT projections below the Reference Margin level of percent to 13 percent by 2022, and continuing to decline to 9.4 percent by Comparing last year s data in Figure 1.10, the 2016 LTRA reference case indicates that ERCOT is now not projected to fall below its Reference Margin Level within the ten-year assessment. This is due to a decrease in net internal demand forecasted in future years and an increase in planned capacity. Comparing forecasted peak load by the year 2025, ERCOT is showing a decrease of 1.2 GW, or a 1.5 percent reduction, from last year s assessment. Similarly, ERCOT is showing an increase in their anticipated resources by 3.8 GW or 4.6 percent increase when compared to last year. Figure 1.10: Texas RE-ERCOT 2015 LTRA and 2016 LTRA Reserve Margin Comparison Figure 1.11 shows the results of a sensitivity analysis whereby all 6.9 GW of unconfirmed coal and natural gas retirements are included in the Anticipated Reserve Margin. From this unconfirmed retirement scenario, ERCOT would fall below their Reference Margin Level of percent to 11.1 percent in 2021 and continue to decline to 5.8 percent by the end of the assessment period. The Prospective Reserve Margin indicates that there are sufficient Tier 2 generation in the queue that may need to be advanced to mitigate a capacity concern. 13 ISONE- Natural Gas Infrastructure Constraints ; September 28,

19 Chapter 1: Reliability Issues Figure 1.11: Texas RE-ERCOT Reserve Margins with Unconfirmed Retirements Over the next ten years, installed resources in the ERCOT Region grow from 16.3 to 26.9 GW nameplate when accounting for Tier 1 planned resources. This would increase the percentage of installed nameplate wind to a total nameplate capacity from 16.5 percent to 22.6 percent. When adding wind and solar, as shown in Figure 1.12, variable resources are projected to be 25.3 percent of the anticipated resource capacity and 42.1 percent of the prospective resource capacity. Actual wind and solar penetration at any time of the year are dependent on weather, irradiance, and resource controllability. Very high penetration levels of variable energy resources (VERs) increases the need for ERSs to effectively dispatch conventional generating units and maintain system reliability. Figure 1.12: Maximum VER Penetration in ERCOT for the 2026 Summer Peak 10

20 Chapter 1: Reliability Issues Figure 1.13 shows the total anticipated capacity between years 2017 and 2026 for ERCOT by fuel type. Total expected net changes across the summer peak include 6.0 GW of natural-gas-fired generation and 0.8 GW of utility-scale VER additions. Figure 1.13: ERCOT Total Anticipated Capacity by Fuel Type In August of 2016, ERCOT set multiple new hourly peak demand records based on preliminary data, settling on 71,197 MW on August 11 th between 4:00 p.m. and 5:00 p.m. 14 At the time of this peak, 4,783 MW of wind was generating on the system, or approximately 6.72 percent of total energy over the hour. 15 System operations indicate that the system is currently capable of delivering energy generated from these wind turbines throughout the assessment area. For parts of the BPS that experience localized events where generation is over-producing in transmission constrained areas, operators must be able to control wind and solar production to prevent contingencies from overloading transmission lines. Similar to the Department of Energy study for MISO, wind curtailments were observed in ERCOT between 2007 and 2015, as shown in their August 2016 report. 16 In December of 2015, Congress passed the Consolidated Appropriations Act, which extended federal renewable electricity production tax credits through the end of These tax credits may encourage new renewable energy development in Texas before sufficient transmission is added that is necessary to effectively deliver this new energy to system load. ERCOT could also experience increased retirements of some fossil fuel generation due to the expected EPA s final regional haze rule 18, 19 and, if upheld by the courts, the potential impacts of the Clean Power Plan (CPP). An initial study by ERCOT also concluded that the Regional Haze requirement would have a significant local and regional impact on the reliability of the ERCOT transmission system. 20 Both of these regulations have been stayed at this time, 21 which creates some uncertainty around the contents of their final ruling. These are currently subject to change. 14 ERCOT Bulletin: ERCOT Breaks Peak Record Again, Tops 71,000MW for First Time"; August 11th, ERCOT Wind Integration Reports 16 U.S. Department of Energy: 2015 Wind Technologies Market Report: Summary; August Renewable Electricity Production Tax Credit Program Info; Energy.gov 18 EPA: Visibility - Regional Haze Program 19 EPA: Air Issues in Texas 20 ERCOT Presentation: Transmission Impact of the Regional Haze Environmental Regulation; October 15, U.S. Court of Appeals for the 5th Circuit: State of Texas et al. v. U.S. Environmental Protection Agency et al., July 15,

21 Chapter 1: Reliability Issues WECC-CAMX Similar to NPCC-New England, there are no risks identified to resource adequacy when only applying a reserve margin analysis. The Anticipated Reserve Margins, shown in Figure 1.14, remain above the Reference Margin Level for all years of the assessment, indicating that there are sufficient resources anticipated to be available to serve peak load. Figure 1.14: WECC-CAMX 2015 LTRA and 2016 LTRA Reserve Margin Comparison However, once additional emerging and standing issues are incorporated into the overall resource adequacy risk analysis, the 2016 LTRA identifies WECC-CAMX as having a medium risk to resource adequacy. This is primarily due to a high reliance on natural-gas-fired generation, limited dual-fuel capability, and the potential reduction in adequate ERSs. Additional study into WECC-CAMX s ramping sufficiency concerns are presented in the Essential Reliability Services Chapter 3. 12

22 Single Fuel Dependency Natural gas continues to be the predominant fuel type in many assessment areas. The tenyear projection for natural gas continues to show an upward slope in the amount of naturalgas-fired resources coming onto the grid, as well as the rate for new-natural-gas fired resources to enter the BPS. Key drivers for this increase include regional initiatives; state renewable portfolio standards; past and potential future regulatory rulings, such as MATS and the CPP; and recent shale gas production. This results in historically low natural gas prices. Chapter 1: Reliability Issues Spot natural gas prices have declined from roughly $13/MMBtu in 2008 to below $3/MMBtu in 2016 as shown in Figure Figure 1.15: Natural Gas Spot Prices Henry Hub This decline in prices has been a large factor in the increase in natural-gas-fired generation. Outside of the effects from regulatory rulings, the low price of natural gas has resulted in additional coal units being shut down as well as the announcement of a significant number of nuclear retirements. Due to historically low gas prices, natural gas use for electricity generation has developed a strong market incentive over all other fuel types within the ERO footprint. The U.S. Energy Information Administration (EIA) routinely monitors monthly energy usage and forecasts. Figure 1.16 shows natural gas usage surpassing coal for the first time in 2015 and for a majority of the time in Figure 1.16: Monthly Net Electricity Generation by Fuel Type 22 EIA Henry Hub Natural Gas Spot Prices 23 EIA Short-Term Energy Outlook - July

23 Chapter 1: Reliability Issues NERC continues to monitor and report on changes to the resource mix and the potential reliability risks associated with these changes. Several ongoing trends have been identified in past LTRAs, such as increasing natural gas, increasing wind and solar, decreasing coal, and uncertainty around the future of nuclear. These previously identified trends indicate an under-forecasted rate of change from assessment to assessment; these rates of change similarly increased in the 2016 LTRA reference case. Figure 1.17 shows natural-gasfired generation data from the 2008 LTRA through the 2016 LTRA reference cases. While the proximity of anticipated natural gas forecasts for year two of each assessment is close to the actual amount installed, there continues to be wide margins between the outer years of each assessment. For example, in 2024 (year ten of the 2014 LTRA and year eight of the 2016 LTRA) there was a forecasted increase of 29.8 GW of net anticipated natural gas; this net change includes both Tier 1 designated planned additions and confirmed retirements. Figure 1.17: Anticipated Natural Gas Capacity by LTRA Reporting Year Figure 1.18: Anticipated Coal Capacity by LTRA Reporting Year The rate of change of retirements of coal plants have continuously increased from assessment to assessment. As shown in Figure 1.18, coal-fired generation data from the 2008 LTRA through the 2016 LTRA reference cases show consistent marginal gaps between assessments. For example, by 2024 of the 2014 LTRA and 2016 LTRA, there was a decrease in the system-wide forecast by 8.5 GW. This trend indicates that coal generation retirements have and are continuing to outpace retirement projections. 14

24 Chapter 1: Reliability Issues Table 1.1 shows that a growing number of the assessment areas are trending towards an increasing dependency on this single fuel source. The table shows a breakdown by assessment area for natural-gas-fired capacity as a percentage of the area s total anticipated capacity. NERC has identified a dependency on a single fuel as a potential reliability risk requiring mitigation. Natural gas has crossed over 50 percent of peak capacity in several areas amidst continued historically low prices and regulatory rulings, which continue to promote increased natural gas generation. Capacity Table 1.1: Natural Gas Percentage of Peak Season Total Anticipated Capacity 2017 (MW) 2021 (MW) 2017 Gas of Total Capacity (%) 2021 Gas of Total Capacity (%) FRCC 35,583 39, % 69.05% WECC-CAMX 40,299 42, % 68.23% Texas RE-ERCOT 45,842 51, % 63.26% NPCC-New England 14,331 16, % 52.33% WECC-SRSG 16,530 16, % 51.84% WECC-AB 8,514 8, % 51.79% SERC-SE 30,256 30, % 46.88% MRO-SaskPower 1,835 2, % 43.97% SPP 30,413 29, % 45.22% SERC-N 19,250 21, % 40.68% MISO 59,566 60, % 42.26% NPCC-New York 16,030 16, % 41.98% PJM 66,760 76, % 38.71% WECC-RMRG 6,695 6, % 38.51% WECC-NWPP-US 20,860 20, % 34.80% SERC-E 15,762 17, % 32.25% NPCC-Ontario 6,568 7, % 24.91% NPCC-Maritimes % 12.66% MRO-Manitoba Hydro % 6.33% WECC-BC % 3.48% NPCC-Québec % 1.33% 15

25 Chapter 1: Reliability Issues Figure 1.19 shows the aggregated summer capacity heat map for all operating and planned natural gas generating units within the ERO footprint. This figure clearly shows that large pockets of natural gas generation are occurring throughout the footprint and evenly spread out. This causes some areas to potentially be more reliant on this single fuel type. Figure 1.19: Natural Gas Generating Units MW Summer Capacity Heat Map NERC conducted a short-term special assessment in 2016 that included an operational risk analysis using NERC s Generation Availability Data System (GADS) database to project natural-gas-fired outages. 24 Using a deterministic approach, all evaluated assessment areas showed no concerns in meeting reserve margin requirements as a result of this operational risk assessment. However, when a single point of disruption, such as unavailable storage facilities, pipeline rupture, or other gas infrastructure failure was considered, reserve margins were jeopardized in some areas. For example, the Aliso Canyon outage in Southern California illustrates the effects of a potential single point of disruption. This one underground gas storage facility in SoCal Gas service territory contains 86 BCF of gas capacity, providing fuel to approximately 9,800 MWs of electric generation. The facility also supports ramping requirements to accommodate the variability of renewable energy resources. This outage has the potential to cause rolling black outs in Southern California until the facility is completely operational again or other mitigation approaches have been employed. 24 NERC: Operational Risk Assessment with High Penetration of Natural Gas-Fired Generation; May

26 Chapter 1: Reliability Issues Gas storage facilities have historically had clear delineations between a summer injection season and winter withdrawal season. Figure 1.20 shows the daily total injections and withdrawals to and from underground gas storage facilities from April August Multiple days within this time period have seen net changes that resulted in more withdrawals than injections. While this is not the first time that system-wide underground storage facilities saw a net withdrawal during the summer season, any continuing changes to use trends for natural gas storage inventories should be monitored and evaluated for potential impacts to future gas availability. Figure 1.20: 2016 Weekly Natural Gas Inventory Changes Many factors must be considered when assessing the potential for reliability risks in resource planning, as outlined in TPL Each assessment area is comprised of a unique set of variables that include existing resources, electric transmission, and natural gas infrastructure. NERC has identified areas that are increasingly reliant on a single fuel type, which increases vulnerabilities, particularly during extreme weather events and conditions. Over the past decade, several areas have significantly increased their dependence on natural gas. This trend has continued amidst historically low natural gas prices and regulatory rulings that continue to promote increased natural gas generation. The assessment identifies four assessment areas with high penetrations of natural gas generation and therefore increased risk through single-fuel dependency: Texas RE-ERCOT, FRCC, NPCC-New England, and WECC-CAMX. Table 1.2 provides a summary of various independent factors that influence these four assessment areas capability to mitigate an increasing reliability risk. 25 NERC TPL

27 Chapter 1: Reliability Issues Table 1.2: Single Fuel Dependency Risk Summaries Alternative Fuel Capabilities Evaluate capabilities across generator fleet, maintain back-up fuel inventories at key stations, and annually test fuelswitching capability Market and Regulatory Rules Provide additional incentives for behavior and investments that support reliability and resiliency Texas RE-ERCOT FRCC NPCC-New England WECC-CAMX Factors that Reduce Risk 14% of gas-fired 73% of gas-fired 30% of gas-fired 4% of gas-fired generation is generation is generation is generation is capable of using capable of using capable of using capable of using alternative fuel. alternative fuel. alternative fuel. alternative fuel. No requirement to maintain back-up fuel inventory Testing of fuel switching is not required Annual winterization and cold weather preparation workshops share lessons learned and best practices to improve reliability during extreme cold weather. No regulatory or market rules exist for maintaining dual-fuel capability and/or firm natural gas transportation. Back-up fuel inventories are required by the state public utility commission. Regulatory rules exist for maintaining dualfuel capability and firm natural gas transportation services and contracts. Winter Reliability Program (through 2018) provides payments for adding dual-fuel capability, securing fuel inventory, and testing fuelswitching capability; compensation for any unused fuel inventory. Recent market rule changes include energy market offer flexibility, timing adjustments to Day-Ahead Energy Market, Winter Fuel Reliability Program, and Payfor-Performance (which starts in June 2018) By creating incentives, the market may indirectly provide incentives for the development of onsite oil, LNG fuel storage, or expanded gas pipeline infrastructure. Very limited backup fuel kept in inventory; holding tanks have largely been removed. No regulatory or market rules exist for maintaining dual-fuel capability and/or firm natural gas transportation within CAISO. 18

28 Chapter 1: Reliability Issues Single Points of Disruption Assess reliability under extreme conditions, loss of major pipeline infrastructure or supply Pipeline Expansion Keep pace with generation expansion and increasing electricity production Limited Exposure to Supply Chain Failure Increase resiliency by maintaining alternative supply chains and paths Meshed pipeline infrastructure significantly reduces this risk No signs of concern with lagging pipeline expansion Natural gas generation is coming on-line with Firm transportation service. Robust supply sources within service area; conventional, shale, and gulf sources The fleet of dual fuel capable generation and ensuring sufficient fuel inventory reduces this risk. Two projects to be completed in 2017 reduce this risk: Sabal Trail Transmission and Florida Southeast Connection. Dual-fuel capabilities reduce the risk Gas supplied from conventional, shale, and gulf sources and transported to Florida; potential LNG import. No local production Area is situated in bottlenecked and physical end of the interstate pipeline system. Two major interstate pipelines connected from southeast; one from the northwest. State policies and pressures have not led to the construction of natural gas pipelines Recent suspension of pipeline projects aimed to support electric generation Gas supplied from conventional, shale, and gulf sources and transported to New England; limited LNG import (supplies Mystic Generation Station~2,000 MW) Area is situated in bottlenecked and physical end of the interstate pipeline system. Two major interstate pipelines connected from southeast; one from the northeast. Aliso Canyon Storage Facility outage continues to impact fuel deliveries. No interstate transmission projects that increase pipeline capacity approved by FERC since Intra-state projects are more likely; however, political opposition continues to challenge expansion (such as in the SoCal Gas North-South Project) Gas supplied from conventional, shale, and gulf sources and transported to California. Some local production No local production; no storage 19

29 Chapter 1: Reliability Issues Maintaining Situational Awareness Electric system operators need awareness of pipeline conditions and must be able to predict generators that may become unavailable Risks Communicated to Policymakers Results and conclusions of studies that evaluate electric reliability should be shared and clarified with state, federal, and provincial policymakers and regulators Maintain Fuel Diversity Maintaining fuel diversity provides inherent resiliency to common-mode risk Pre-season surveys of fuel inventories Coordination with pipeline operators Gas Curtailment Risk Study (2012) Annual LTRA NERC will evaluate single points of disruption in a 2017 special assessment. 37% Gas 63% Non-Gas No significant changes Annual LTRA NERC will evaluate single points of disruption in a 2017 special assessment. 31% Gas 69% Non-Gas Pre-season surveys of fuel inventories Improved coordination and information-sharing with natural gas pipeline operators Coordination of generator and pipeline maintenance schedules. Gas Usage Tool estimates the remaining gas pipeline capacity by individual pipe for use by ISO-NE system operators Annual LTRA Key participant in the Eastern Interconnection Planning Collaborative (EIPC) gas-electric interface study NERC will evaluate single points of disruption in a 2017 special assessment. 48% Gas 52% Non-Gas Improved coordination and information-sharing with natural gas pipeline operators Joint Agency Daily Reliability Communication (throughout Aliso Canyon outage) Development of gas curtailment methodology and scenario planning Active coordination on energy emergencies with California Energy Commission; action plans to respond to Aliso Canyon Storage Facility outage. Annual LTRA NERC will evaluate single points of disruption in a 2017 special assessment. WECC is working with the natural gas industry to study potential impacts to reliability as the Western Interconnection becomes more reliant on naturalgas-fired generation. 32% Gas 68% Non-Gas 20

30 Chapter 1: Reliability Issues Recognizing the increased dependence on natural gas as a fuel for electricity generation, FERC has already taken steps to improve the coordination of wholesale natural gas and electricity market scheduling: FERC Order No. 787, issued on November 15, 2013 (Rulemaking RM ), 26 provides explicit authority to interstate natural gas pipelines and public electric utilities participating in the interstate commerce to share nonpublic operational information with each other to promote reliable service or operational planning on their systems. FERC Order 809, issued on April 16, 2015 (Rulemaking RM ), 27 provides for better coordination of the scheduling practices of the wholesale natural gas and electric industries, as well as additional contracting flexibility to firm natural gas transportation customers through the use of multi-party transportation contracts. However, regulatory and policy solutions that help expand pipeline access, reliability, and the needs of electric generation have a not surfaced. The recent suspension of Kinder Morgan s AED and Algonquin s proposal to facilitate electric utility purchase of pipeline capacity demonstrates the need for regulatory solutions to facilitate electric generator commitments. This is particularly true for generation operating in wholesale electric markets. Recommendations As natural-gas-fired resources continue to increase, system planners and operators should evaluate the potential effects of an increased reliance on natural gas as it pertains to BPS reliability. Natural gas provides just-in-time fuel; therefore, firm transportation and maintaining dual-fuel capability can significantly reduce the risk of common-mode failure and wider-spread reliability challenges. As part of future transmission and resource planning studies, planning entities will need to more fully understand how impacts to the natural gas transportation system can impact electric reliability. Regulatory action may be needed to better calibrate electric and gas industries. 26 FERC Order 787; U.S. Docket No. RM FERC Order 809; U.S. Docket No. RM

31 Chapter 1: Reliability Issues Nuclear Uncertainty Lower natural gas prices driven by abundant domestic supply, along with other economic and regulatory factors, have pressured the economic viability of 99 operable nuclear units. Confirmed and unconfirmed retirements of facilities are projected in California (Diablo Canyon), Illinois (Quad Cities and Clinton), Massachusetts (Pilgrim), and Nebraska (Fort Calhoun) during the next ten years. Other at-risk units are located in the northeastern states. Uncertainties and contributing factors to nuclear retirements include high operating costs with low prevailing power prices, regulatory issues, and public opposition. 28 Figure 1.21 shows the forecasted and potentially advanced total nuclear capacity that could be affected. It is projected that 6.4 GW of capacity is ready to retire; this makes up five percent of the total installed capacity by The High Nuclear Retirement Case capacity values from NERC s Clean Power Plan; Phase II Assessment 29 indicate that a potential total of 26.8 GW, or 21 percent of all anticipated nuclear capacity, could be at risk to retire under this scenario. Figure 1.21: Anticipated Nuclear Capacity by 2026 New York regulators recently introduced an energy plan that will preserve the economic viability of three upstate units. However, the retirement of seven remaining at-risk units would continue a recent trend since 2012 of decommissioned units in Florida (Crystal River), Wisconsin (Kewaunee), California (San Onofre), and Vermont (Vermont Yankee). Despite economic pressures on existing units, the Tennessee Valley Authority (TVA) started the Watts Bar two unit in mid-2016, following nearly three decades of a sluggish pace for new unit builds. According to the 2016 LTRA Reference Case, four additional units are expected to be completed by These additions, combined with ongoing uprates to existing units in the United States, will result in a continued steady contribution of nuclear power over the next ten years. Similar legislation was passed by the Illinois General Assembly that would authorize subsidies totaling $2.4 billion over the next decade to allow both Quad Cities and Clinton nuclear facilities to remain open. 30 Recommendations NERC should continue to monitor the potential effects of nuclear retirements on overall resource adequacy as well as potential mitigating factors such as state regulatory measures that provide incentives for nuclear facilities to remain operational. 28 World Nuclear: Nuclear Power in the USA; September 26, Potential Reliability Impacts of EPA's Clean Power Plan Phase II; May Illinois General Assembly: Bill Status of SB

32 Chapter 2: Probabilistic Analysis Probabilistic analyses describe events in terms of how probable they are and include performance characteristics of BPS components, such as generator outage rates, resource realizations in terms of energy produced, load characteristics, and transmission congestions or constraints. A prediction of future reliability must be expressed in terms of the expected performance of the system components and the uncertainty in those expectations. Probabilistic methods typically rely on either statistical analyses of historical performance or enumeration techniques that are capable of simulating large numbers of contingencies. However, the choice of methods and selection of acceptable reliability levels are still matters of judgment and differ from Region to Region (and from utility to utility in some cases). The analytical processes used by resource planners range from relatively simple calculations of planning reserve margins to rigorous reliability simulations that calculate system Loss of Load Expectation (LOLE) or Loss of Load Probability (LOLP) values. 31 The one-event-in-ten-year (0.1 events per year) LOLE is produced from this type of probabilistic analysis. This planning criterion requires an electric system to maintain sufficient capacity such that system peak load is not likely to exceed available supply more than once in a ten-year period. Utilities, system operators, and regulators across North America rely on variations of the one-event-in-ten year criterion for ensuring and maintaining resource adequacy. Sensitivity Model Sensitivity analyses around Monto-Carlo-simulated reserve margins can be run by treating specific (independent) variables as random. To demonstrate how any independent variables impact reserve margin results, a simulation was run for a generic summer-peaking system. Figure 2.1 shows the resulting reserve margin uncertainties from 2018 to 2026 using probabilistic distributions of wind and solar power from actual time series profiles. Due to the random variability of the simulated wind and solar, the uncertainty around the calculated reserve margin mean is demonstrated. This is indicated in the figure whereby the area in blue shows the resulting bandwidth of one standard deviation from the mean and the grey shows two standard deviations from the mean. Figure 2.1: Reserve Margin Uncertainty Due to Wind and Solar Variability 31 A traditional planning criterion used by some resource planners or load-serving entities is maintaining system LOLE below one day in ten years. Loss-of-Load Expectation (LOLE) is generally defined as the expected number of days per year for which the available generation capacity is insufficient to serve the daily peak demand. This is the original metric that is calculated using only the peak load of the day (or the daily peak variation curve). However, this metric is not being reported as part of this assessment. Currently, some Assessment Areas also calculate the LOLE as the expected number of days per year when the available generation capacity is insufficient to serve the daily load demand (instead of the daily peak load) at least once during that day. 23

33 Chapter 2: Probabilistic Analysis This generic model was further explored by running additional simulations to other applied independent variables. DSM, Tier 1 capacity resources, and firm transactions risk profiles are developed based on historical statistical performance data provided in the NERC 2015 Electricity Supply & Demand database and the use of engineering judgments. Figure 2.2 shows a tornado diagram highlighting the sensitivity of these other input parameters and ranking them by their effect on reserve margin mean values for The figure ranks the sensitivity of an output reserve margin from different independent variables. The following variables were considered for this analysis: demand, wind, solar, demand-side management (DSM), existing-certain resources, Tier 1 resources, and firm capacity imports and exports. Figure 2.2 shows how much a one standard deviation change in input variables affects the output reserve margins. The width of each parameter directly represents the impact an independent variable can have on reserve margins and the degree of which uncertainty can be assumed. Figure 2.2: Tornado Plot for 2016 Simulated Reserve Margin Ranked by Effect of the Reserve Margin s Mean Probabilistic distributions were assigned to hourly demand, wind, and solar power profiles. Additionally, risk profiles were applied to DSM, Tier 1 capacity resources, and firm transactions based on past performances. The results in Figure 2.2 shows that demand is the most impactful parameter that drives this generic system s simulated reserve margin. Specifically, accurate load modeling and forecast uncertainty modeling are critical aspects for effective resource adequacy planning. The second most sensitive parameter is wind power due to the large percent share of the system s generation mix. This analysis also shows the least sensitive and least influential input parameter is solar due its relatively small percent share of the generation mix. As each system includes a unique set of independent variables, and this type of analysis is helpful in determining the most significant parameters of a given system. Planners can then judge which risks to take and which ones to avoid in maintaining resource adequacy while allowing for best operational and planning decisions under prevalent uncertainty. 24

34 Chapter 2: Probabilistic Analysis VER Capacity Contributions For a reserve margin analysis, the capacity contributions of installed variable energy resources (VERs) are the values that are expected to be available to an assessment area across the peak load hour. The calculation of the capacity contribution of conventional generating units are straightforward and are based on unit performance ratings, forced outage rates, and annual unforced maintenance cycles. However, the capacity contributions of VERs are not intuitive due to their inherent characteristics. There are two major attributes of variable generation that notably impact bulk power system planning and operations: 32 Variability: The output of variable generation changes according to the availability of the primary fuel (e.g., wind, sunlight, and moving water), resulting in fluctuations in the plant output on all time scales. Uncertainty: The magnitude and timing of variable generation output is less predictable than for conventional generation. Many factors are affecting the system-wide increases of renewable resources and are the predominant choice for renewable energy integration. In high-ver penetration conditions, a larger portion of the total resource portfolio will be comprised of energy-limited resources when applied to today s power system. This fact somewhat complicates, but does not fundamentally change, existing resource adequacy planning processes as they are still driven by a reliability-based set of metrics. Resource adequacy can be confirmed through detailed reliability simulations that compare expected demand profiles with specific generating unit s forced outage rates and maintenance schedules to yield LOLE or LOLP values. Reliability simulations typically include probabilistic production cost simulations for meeting a specified demand curve (or chronological curve) from a specified generation fleet while incorporating the forced and unforced outage rates over the simulation period. Current approaches used by resource planners fall into four basic categories: 33 A rigorous LOLE/LOLP-based calculation of the effective load-carrying capability (ELCC) of variable generation relative to a benchmark conventional unit Calculation of the capacity factor of the variable generation during specified time periods that represent high-risk reliability periods (typically peak hours) A tailored approach for applying a historical performance rolling average (typically 2 3 year) Applications based on policies established through a nontechnical analysis 32 More details on variable generation attributes can be found on the now-disbanded Integration of Variable Generation Task Force Special report on Standard Models for Variable Generation 33 More information of these approaches can be found in NERC s Special Reliability Assessment: Methods to Model and Calculate Capacity Contributions of Variable Generation for Resource Adequacy Planning 25

35 Chapter 2: Probabilistic Analysis The capacity contribution component of VERs differs greatly from that of conventional generation. Conventional generation uses typical summer and winter ratings that do not differ greatly from the nameplate capacity rating. Because the capacity contributions from VERs are a statistical representation of normal operations, NERC monitors the methods and assumptions for calculating these components. In addition to assessment areas using varying methods to calculate capacity contributions for future generation, additional variances arise when considering areas with capacity market structures. Figure 2.3 and Figure 2.4 show the wind and solar nameplate values and their on-peak capacity contributions anticipated by all applicable assessment areas for Figure 2.3: 2026 Existing and Tier 1 Wind Figure 2.4: 2026 Existing and Tier 1 Solar 26

36 Chapter 2: Probabilistic Analysis A generic summer-peaking model was further explored to analyze any changes to reserve margin requirements due to changes in VER capacity contribution levels. This additional analysis maintained an equal quantity of all resource types but applied a high and low capacity contribution for both solar and wind units. Figure 2.5 shows that for higher VER capacity contributions, the Reference Margin Level would need to be increased to maintain a constant LOLE of 0.1 days per year. If capacity contributions are calculated accurately or conservatively, then the current Reference Margin Level would be sufficient. Figure 2.5: Reference Margin Level Increases with Added VER Due to the identified relationship between Reference Margin Level and VER capacity contributions, consistent and accurate methods are needed. There are existing simplified approaches to calculate VER capacity values, and these can be easily extended to cover other forms of variable generation. In general, these methods calculate the resource s capacity factor over a time period that corresponds to system peaks. These approaches can provide a reasonable, simple approximation for capacity values. However, system characteristics, in some cases, may result in a mismatch between a rigorously calculated ELCC and a peak-period capacity factor as an approximation of capacity value. Simplified approaches should be benchmarked and calibrated to the rigorous ELCC calculations to ensure the validity of any approximation. 27

37 Chapter 3: Essential Reliability Services The North American electric grid is experiencing a shift in the resource mix, driven by a variety of factors that include retirements of conventional resources and the integration of new resources. This leads to potential impacts on essential reliability services (ERSs), such as frequency, voltage, and ramping capability. This transformation in the resource mix will change the planning and operation practices of the current electric grid. Although many resources are able to provide the essential services needed to maintain BPS reliability, understanding system characteristics and related behaviors will aide in successful integration of new technologies. BPS planning and operations will be tailored to incorporate this transformation in order to maintain reliability. In order to study these implications, NERC formed the Essential Reliability Services Task Force (ERSTF), which produced its final report in December The report studied the three reliability blocks: 1) Frequency Support, 2) Voltage Support, and 3) Net Demand Ramping Variability. The task force developed a total of nine measures for the essential services, conducted data analyses for five of them using three years of historical data and three years of forward looking data, and proposed considerations for industry practices for the remaining measures. In addition, the task force studied the potential impact of a substantial penetration of distributed energy resources (DERs) that, in aggregate, could impact the reliability of the BPS. Finally, the report recommended the developed measures be continually monitored for any trends that could potentially impact reliability. A summary of the recommendations from the framework report are listed below: All new resources should have the capability to support voltage and frequency. Ensuring that these capabilities are present in the future resource mix is prudent and necessary. The measures are intended to highlight aspects of reliability that could suggest future reliability concerns. They should be addressed with suitable planning and engineering practices. Planning and operating entities should use industry practices that will help ensure that emerging concerns are addressed with system specific planning and engineering practices. The task force recognized that DERs will increasingly affect the net distribution load that is observed by the BPS. Pursuant with NERC s reliability assessment obligations, the ERSTF further recommends that NERC establish a working group to examine the forecasting, visibility, control, and participation of DERs as an active part of the BPS. With prudent planning, operating and engineering practices, and policy oriented to support reliability, DERs should be able to be reliably integrated into BPS operation. 35 The reliability of the system can be maintained or improved as the resource mix evolves, provided that sufficient amounts of ERSs are available. This can be achieved by sharing of experiences and lessons learned around the industry. In 2016, the ERSTF transitioned to a working group (now known as ERSWG) and was charged with examining methodologies to determine sufficient levels of each ERS. In addition, this working group was asked to form a task force under their purview to address challenges and potential risks from increasing DERs. The ERSWG is presently formulating a whitepaper that focuses on methods to develop sufficiency guidelines around the proposed ERSTF measures. These sufficiency guidelines are more process oriented and include the following: frequency response, voltage limits, and ramping models that tend to vary by particular area and Balancing Authority. The whitepaper further explores important technical considerations so the industry can understand, evaluate, and prepare for the increased deployment of variable energy resources (VERs), retirements of conventional coal units, increases in demand response (DR) and distributed technologies, and other changes to the traditional characteristics of generation and load resources. The working group is currently in the process of 34 ERSTF Final Measures Framework Report 35 The Distributed Energy Resources Task Force (DERTF) was created in early

38 Chapter 3: Essential Reliability Services collecting data for the proposed measures and evaluating them by using the sufficiency methodology, the results of which will be finalized by end of Frequency Support Frequency support is provided through the combined interactions of synchronous inertia, primary frequency response (PFR), and secondary frequency response. Working in a coordinated fashion, these characteristics and services arrest the decline in frequency and eventually return the frequency to the desired level. Figure 3.1 below shows a typical frequency excursion and recovery, whereby the red line indicates the initial rate of change of frequency (RoCoF) due to inertial responses from synchronous machines. Figure 3.1: Typical Frequency Excursion and Recovery It is important to determine various levels of inertia for a future resource mix to ensure that the system doesn t fall below a minimum level. Some newer technologies, such as wind turbines, provide the capability to inject real power at a fast rate during a frequency excursion, and thus all resources need to be taken into account in planning and operating considerations of a system. With the increasing use of nonsynchronous generation and other electronically-coupled resources (both generators and loads), the level of synchronous inertial response is reduced. This leads to a need to consider both the amounts of synchronous inertia and the available amounts of PFR based on expected conditions. Frequency support encompasses inertia, nadir, PFR, and secondary frequency response. While inertia is just a component of overall frequency response, it plays an important role in arresting the RoCoF and prevents the nadir from reaching the level of under-frequency load shedding. Measure 4 evaluates the detailed anatomy of a frequency excursion, such as adding the calculation of Point C and time parameters in a typical event as shown in Figure 3.1. Measure 4 will be analyzed further in the 2017 State of Reliability Report for analysis of qualified 2016 frequency events. Measures 1 3 in the ERSTF Framework Report analyze the inertia and associated RoCoF. The results of data gathering for inertia and RoCoF measures for ERCOT and the Eastern Interconnection (EI) are presented here. 29

39 Chapter 3: Essential Reliability Services ERCOT In Texas, ERCOT determined it is important to track system-level inertia in real-time to ensure system reliability. Figure 3.2 shows the snapshot of ERCOT s real-time dashboard for inertia. The operator is able to observe total inertia and load in this single display. The dashboard also shows 24-hour inertia contributions by generator type. Monitoring by types of resources is done to enable more granular analysis of inertia trends in real-time. Figure 3.2: ERCOT Dashboard to monitor real-time inertia As part of the trend monitoring of ERSWG Measures 1-3, ERCOT (and other interconnections) started collecting the inertia data on June 1, Figure 3.3 represents the inertia data for ERCOT by hour in box plot format. On each box, the central mark (red line) is the median, the edges of the box (in blue) are the 25th and 75th percentiles, the whiskers correspond to +/- 2.7 sigma (i.e., represent 99.3 percent coverage, assuming the data are normally distributed), and the outliers are plotted individually (red crosses). 30

40 Chapter 3: Essential Reliability Services Figure 3.3: ERCOT Inertia by Hour of Day (June 1 July ) Figure 3.4 represents a more granular version of Figure 3.3 in that it shows ERCOT s inertia data in 15-minute intervals. Ultimately, both figures display the pattern of system inertia that mostly coincides with load levels. Inertia at low load levels becomes a challenge for providing frequency response in case of a frequency excursion. Figure 3.4: ERCOT Inertia by 15-minute Interval of Day (June 1 July ) 31

41 Chapter 3: Essential Reliability Services Eastern Interconnection The Eastern Interconnection (EI) has significant levels of synchronous generation that allow sufficient contributions of inertia as shown in Figure 3.5 below. The figure shows the total interconnection inertia contributions following changes to load levels. Based on the availability of data, system level monitoring of inertia and thus potential frequency response can be evaluated. Figure 3.5: Eastern Interconnection System Inertia Primary Frequency Response Analysis PFR, as shown in Figure 3.1, has been identified by NERC s Operating Committee and Planning Committee as an ERS that will be affected by the changing grid characteristics of the North American BPS. The changing grid will be characterized by an increased penetration of new technology resources and the retirement of conventional generating resources. NERC is studying PFRs, which relate the size of the resource lost to the resulting net change in system frequency during the period when stabilizing frequency is determined following the initiating event. To study the changing characteristics and PFR performance of the grid, NERC will use power system computer planning models. These planning models will be used to develop scenarios of the future system that will be studied to gain a detailed understanding of various factors affecting PFRs and the resulting under-frequency load shedding (UFLS) system operation. UFLS systems are designed as a backstop to prevent such events from cascading across the BPS. Primary frequency controls are deemed adequate if, following the sudden loss of the largest generator, the primary frequency control response provided by on-line resources successfully arrests and stabilizes frequency decline. This should be prior to initiating any UFLS action to arrest further frequency decline by dropping firm customer loads Largest generation loss is defined as largest category C (N-2) event; except for Eastern Interconnection, which uses largest event in the last 10 years. 32

42 Chapter 3: Essential Reliability Services NERC will perform various analyses to study the effect of replacing conventional generation with increased penetrations of new technology resources. These analyses will determine how various future scenarios of NTR plant additions and Clean Power Plan (CPP) retirements impact the system PFR. With understanding from these analyses, NERC will develop an objective basis for understanding the reliability implications of varying levels of integration and penetrations of NTRs on PFR. This study therefore has three major purposes: Understand the interconnection-wide reliability implications of the changing resource mix on frequency response. Evaluate policy issues, such as the CPP implementation, with sensitivities of the changing resource mix, control strategies, and other assumptions. Support future proposals for rule making. The final report will be an assessment of the reliability risks and it will recommend technical guidance to mitigate risks of encountering system conditions that will result in UFLS protective actions. This report will document the results of the PFR evaluations, provide a firm basis for future system reliability risk determinations, and identify potential solutions that will assure a continuation of an adequate level of frequency response is maintained for the reliable operation of the BPS using conventional and new technology resources. This frequency response study will be completed in three phases over approximately three years. 1. Phase I will study PFR for the EI by using detailed frequency modeling for existing and future plants with a load modeled by the ZIP Load Model. The ZIP Load model is characterized by coefficients of a load model comprised of constant impedance (Z), constant current (I), and constant power loads (P). 2. Phase II will study PFR for the EI by using detailed frequency modeling for existing and future plants with load modeled by a complex load model (CLM). The CLM model will replace the constant MVA, current, and impedance load with a composition of loads consisting of large and small induction motors, discharge lighting, constant MVA load, and a static load response. 3. Phase III will study PFR for the Eastern Interconnection, ERCOT, and the Western Electricity Coordinating Council by using detailed frequency modeling for existing and future plants, loads modeled by a CLM model, DER modeling, and energy storage modeling. Net Load Ramping Changes in the amount of nondispatchable resources, transmission system constraints, load behaviors, and resource mix can impact the ramp rates needed to keep the system load and generation in balance. Most areas that have wide-scale integration of DERs may experience planning or operational issues due to large load ramps. For example, solar photovoltaic (PV) is heavily integrated into the electric system in California ISO (CAISO), and operators there are faced with multi-hour ramps during the evening hours partly coinciding with sunset. Hence, CAISO must ensure that enough system ramping capability is available to follow the fast net load fluctuations. With the increasing penetration of generation resources for which the BA may have limited ability to control the level of output, consideration of system ramping capability becomes an even more important component of planning and operations. CAISO has been experiencing challenges with ramping inter- and intra- hour; these challenges were studied and presented through the duck curve, 37 shown below in Figure 3.6. On May 15, 2016, actual net-load dropped to 11,663 MW from the projected 2016 load of just over 15,000 MW, which is four years ahead of the original duck curve estimate. Thus CAISO is experiencing the issues with ramping in advance of the previously predicted time frame. 37 CAISO - Flexible Resources to Help Renewable - Fast Facts 33

43 Chapter 3: Essential Reliability Services Figure 3.6: Analysis of CAISO Duck Curve Based on the experience with net load ramping, CAISO projects the future ramps to be higher than previously projected. Below are the projections for CAISO s ramping from 2015 actual ramps to 2019 projected one-hour upward and downward ramps, shown in Figure 3.7 and in Figure 3.8 respectively. The three-hour upward and downward ramp projections are shown in Figure 3.9 and Figure 3.10 respectively. 34

44 Chapter 3: Essential Reliability Services Figure 3.7: CAISO Monthly One-Hour Upward Ramps Figure 3.8: CAISO Monthly One-Hour Downward Ramps 35

45 Chapter 3: Essential Reliability Services Figure 3.9: CAISO Monthly Three-Hour Upward Ramps Figure 3.10: CAISO Monthly Three-Hour Downward Ramps 36

46 Chapter 3: Essential Reliability Services ERCOT s Resource Flexibility Analysis Method EPRI has created several flexibility metrics that can be helpful for performing ramping analysis. One metric is periods of flexibility deficit, which is a count of the number of intervals in the study period where net flexibility is below zero. Periods of flexibility deficit is calculated for a specific time horizon as well as for ramp direction. Another metric is expected unserved ramp, which is the total magnitude of negative net flexibility. ERCOT is using the methodology for performing this analysis. The methodology assumes that the study period is a full year and the time resolution is five minutes, but it can be modified to accommodate different study periods and time resolutions. The first step in performing the flexibility study for a future year is to prepare a generator database of all generators that will be active in the study year; this includes existing generators that are not scheduled to retire before the study year as well as planned generators that are scheduled to be on-line during the study year. The next step is to obtain hourly profiles for wind, solar, and load for the study year, and then interpolate these profiles into a higher resolution. The hourly wind, solar, and load profiles are used as inputs into a production cost simulation tool along with the generator database prepared earlier. Any resources with fixed schedules or must-run units should be modeled as such in the production cost simulation. ERCOT uses Energy Exemplar s Plexos to perform 365 daily optimization runs with a one-day look ahead in order to obtain unit commitments for the entire study year. Each unit s commitment status must be found for each hour of the year from the hourly production cost simulation run, and the temporal resolution of the hourly wind, solar, and load profiles must be increased to a five-minute resolution. Once this is due, a sequential production cost simulation should be performed in order to obtain each unit s five-minute dispatch for the study year to serve the given net load. Again, in this dispatch run, capacity reserved for provision of the AS is not available for dispatch except for nonspinning reserves that are generally available in ERCOT in energy scarcity situations. This dispatch data, along with the generator database and the five-minute load data, are the inputs to EPRI s InFLEXion tool. The InFLEXion tool can perform all of the analysis described in this document as well as additional ramping/flexibility analyses. 37

47 Chapter 3: Essential Reliability Services ERCOT s Sample results Figure 3.11 below shows the average one-hour net load ramps for each hour of each day in the same time period. Negative numbers in this figure represent hours where the average net load ramp was in the downward direction rather than upward. Figure 3.11: ERCOT s Net Load One Hour Ramps for Spring 2017 The heat map Figure 3.11 shows the pattern of upward or downward ramps for the spring of 2017 by the hour of the day. The pattern forecasts ramps to occur during the early morning and at late night. In addition to calculating the ramp up, ERCOT forecasted available flexibility resources that can be used to meet the ramps. Figure 3.12 below shows the average amount of available upward flexibility that can be provided by the resource fleet to serve one-hour net load ramps for each hour of each day in the spring of The available flexibility is calculated based on what units are on-line and able to increase generation from their current operation point (as obtained from production cost simulation) as well as units that are off-line but could start up quickly enough to help serve the ramp (if necessary), including off-line nonspinning reserves. Note that in this level of analysis, all reserves are considered available to follow expected net load ramps Figure 3.12: ERCOT s Available Flexibility for One Hour Ramps in Spring

48 Chapter 3: Essential Reliability Services ERCOT is currently working on improving unit commitments to include the tradeoff between using coal and natural gas fuels. The PLEXUS tool is optimal for plant startup and run times, but it doesn t consider human behavior as well as shortened or prolonged maintenance schedules. In addition, EPRI is considering improving the process for their InFLEXion tool to include transmission constraints. Hourly CPS1 Evaluation on Interconnection Basis For areas where ramping is not a significant challenge, there needs to be a different method of evaluating whether that area is or would be experiencing ramping-related challenges. Historical Control Performance Standard (CPS) hourly data can be an indication of potential ramping issues. A BA can evaluate its historical ramps against its realtime control performance standard (CPS1) 38 to determine whether it is beginning to experience ramping problems. This can be accomplished by evaluating hourly CPS performance data for trends, such as CPS1 scores less than 100 percent for certain hours of the day and certain months of the year. It should be noted that this evaluation of CPS1 on an hourly basis does not imply any NERC standards requirements; this is simply a methodology for evaluating ramping needs of a given area. Figure 3.13 below shows the CPS1 score exceedances by BA on an hourly basis for the Eastern Interconnection (EI). The bottom graph (Grey) shows the availability of CPS1 data from the BA, the middle graph (Red) shows number of hourly CPS1 exceedances for three or more consecutive hours, and the top graph (Black) shows the number of hourly CPS1 exceedances. The last column, which is highlighted in a blue box, represents the total score of hourly CPS1 for EI in each respective area. Figure 3.13: Eastern Interconnection CPS1 exceedance counts per BA on an hourly basis 38 CPS1 is a statistical measure of a BA s area control error (ACE) variability in combination with the interconnection frequency error from scheduled frequency. It measures the covariance between the ACE of a BA and the frequency deviation of the interconnection, which is equal to the sum of the ACEs of all of the BAs. CPS1 assigns each BA a share of the responsibility for controlling the interconnection s steadystate frequency. The CPS1 score is reported to NERC on a monthly basis and averaged over a 12-month moving window. A violation of CPS1 occurs whenever a BA s CPS1 score for the 12-month moving window falls below 100 percent. 39

49 Chapter 3: Essential Reliability Services Similarly, Figure 3.14 shows the CPS1 exceedances for Western Interconnection (WI). Note that the WI chart does not include BA s that use variable bias settings to calculate their ACE. Figure 3.14: Western Interconnection CPS1 exceedance counts per BA on an hourly basis 39 Several assessment areas, including MISO, Manitoba Hydro, and NPCC-Maritimes, evaluated their respective area for ramping-related challenges. From their results, the ramping measure continues to be monitored, but does not currently pose a challenge to reliability. 39 Balancing Authorities unnamed to maintain confidentiality. 40

50 Chapter 3: Essential Reliability Services Distributed Energy Resources An increasing quantity of DERs are being installed behind-the-meter and they may or may not be known assets on the distribution system. Figure 3.15 shows the cumulative installations of nonutility scale solar generation in the U.S. from Additional installations are forecasted to Behind-the-meter generation units essentially means that, despite some knowledge of how much potential capacity is installed, there is no individual metering of these units that would indicate their actual energy produced during any time frame. Including generation from these units and the real demand of the system results in a netted system load that can complicate system operations. In low penetrations of DERs to local system load, there is much less potential for any issues to be escalated into the BPS. The potential for issues to impact the BPS grows in some relation with increased DER penetration levels. Both the potential issues to system reliability and the changing characteristics of the load due to increasing DER installations must be studied further. Figure 3.15: Cumulative Installations of Non-Utility Photovoltaic Historical data sets enable characterization and trending of key performance metrics, including factors that contribute to resource availability and adequacy. DERs, such as rooftop solar, do not have long-term historical data sets, and this lack of data limits the understanding of the long-term implications of DER performance. The potential output levels of DERs show a large degree of variance over a vast geographic scale, so the ideal type and capacity contributions of DER generation will differ by region. Several studies for capacity value calculations, however, differ in results due to differences in DERs and load characteristics in the regions under study. Calculating capacity values for existing DERs requires chronological generation data that are synchronized with load data and other relevant system properties. Existing power system data bases can be used to track this data, which would be useful in helping to better understand DER performance and operational issues. However, consistent and accurate methods are needed to calculate capacity credits (sometimes called capacity values) attributable to DERs. Defining a compendium of best practices for evaluating DER contribution to resource adequacy would assist in providing some alignment of these different methods. Data and information exchange across the transmission and distribution interface is a crucial aspect of power system planning, forecasting, and DER modeling. Both transmission and distribution entities should develop a 40 GTM Research: Solar Market Insight Report 2016 Q2 41

51 Chapter 3: Essential Reliability Services common framework where this type of data exchange can be facilitated to ensure the reliability of the BPS. In addition, adequate operator observability and controllability of the BPS will require access to information and data concerning existing and planned DERs. System planners, both transmission and distribution, should be assessing the penetration of large amounts of DERs that may require changes to forecasting, dispatch, and control of the bulk power system. In states where policies haven t yet incentivized the installation of DERs, policy makers should consider the potential reliability threats and incorporate system upgrades into future policy decisions. State policy makers should leverage the experience in California, New York, other states and provinces that continue to refine their respective policies for accommodating high levels of distributed generation in a reliable manner. Diminishing On-Peak Impact of DER At a certain penetration level of distributed or behind-the-meter energy resources, additional installations of photovoltaic (PV) resources have a diminished impact on peak load. This is mainly due to these resources being considered passive load modifiers rather than dispatchable resources, thus netting their generated energy in with the load. A generic summer-peaking system was modeled to study the impacts of increasing DERs; this result is depicted in Figure 3.16, which shows that increased solar penetration can shift the critical hours to later in the day with a largely coincident summer-peaking load profile. Due to lower solar irradiances in late hours of the day, the more solar added to the system, the less significant it becomes at peak demand. This affects the timing of reliability in critical hours and decreases the capacity contributions that can be expected to serve load during critical hours. Figure 3.16 shows net load has been shifted 2 3 hours at high levels of solar penetration (i.e., from 10 percent to 50 percent solar). Figure 3.16: Demand and Net Demand Shapes at Different DER Penetration Levels 42

52 Chapter 3: Essential Reliability Services Furthermore, Figure 3.17 shows that solar capacity factors reach near-zero levels as solar penetration increases. This ultimately supports the results of this system analysis whereby net load is shifted 2 3 hours due to capacity factors reaching near-zero levels and as solar energy radiation in the evening is negligible. Figure 3.17: Solar Power Capacity Factor Distributed Energy Resources Task Force The Distributed Energy Resources Task Force (DERTF) was established in response to a recommendation of the Essential Reliability Services Task Force (ERSTF) Measures Framework Report. 41 This task force will develop a report by Q1 of 2017 that will examine existing practices for incorporating DERs in to planning models and studies, identify operational impacts to the BPS, and review existing NERC standards to ensure that DERs can be integrated reliably into the BPS. The report will also explore existing policies oriented to support the reliable integration of DERs on the BPS and further examine the interplay with other ERSs. In developing this report, the task force will review the NERC Functional Model, existing NERC Reliability Standards, and coordinate with Electrical and Electronics Engineers (IEEE) 1547 related efforts. Additionally, the task force will review definitions for behindthe-meter generation, distributed generation, and other related terms to provide clear distinctions between each category. 41 NERC Essential Reliability Services Task Force Measures Framework Report; November

53 Chapter 4: Reliability Assessment Trends and Analysis This section provides an overview of the key projections collected and analyzed for this assessment, including reserve margins, demand and energy, demand-side management (DSM), generation fuel mix, and transmission adequacy. These data and resulting analyses are crucial components in the assessment and identification of the reliability issues in focus. Reserve Margins Understanding the relation between changes to an area s demand needs, and available generating capacity is traditionally performed through a planning reserve margin analysis. Generally, this analysis compares the forecasted peak load to the amount of capacity that could be considered available to serve peak load. Included in these values are considerations for passive or controllable peak load reduction programs, also referred to as DSM. When compared to an individual area s target reserve margin level (or Reference Margin Level), this deterministicbased calculation provides a straightforward viewpoint on the adequacy of a system s resources for all ten years of the assessment. Both an Anticipated Reserve Margin and Prospective Reserve Margin are calculated using a deterministic reserve margin analysis, essentially providing two different benchmarks for varying degrees of certainty in future generation. More details on the components to both of these reserve margin calculations can be found in Appendix II. Table 4.1, 42 Table 4.2, and Table 4.3 below show the overall demand, resources, and planning reserve margins for Years 1, 5, and 10 of the assessment period for all assessment areas and interconnection subtotals. Table 4.1: Peak Season 2017(S) / (W): Projected Demand, Resources, & Planning Reserve Margins Assessment Area/ Demand (MW) Resources (MW) Reserve Margins (%) Reference Interconnection Total Net Anticipated Prospective Anticipated Prospective Margin Internal Internal Level FRCC 48,125 45,111 55,015 55, % 22.89% 15.00% MISO 127, , , , % 23.78% 15.20% MRO-Manitoba 4,826 4,826 5,419 5, % 14.51% 12.00% Hydro* MRO-SaskPower* 3,724 3,639 4,303 4, % 18.24% 11.00% NPCC-Maritimes* 5,584 5,312 6,716 6, % 26.79% 20.00% NPCC-New England 26,698 25,857 31,112 31, % 21.10% 16.74% NPCC-New York 33,363 32,115 39,613 40, % 25.74% 15.00% NPCC-Ontario 22,680 22,000 26,822 26, % 21.92% 18.13% NPCC-Québec* 38,150 35,982 41,217 42, % 17.61% 12.20% PJM 154, , , , % 33.95% 16.50% SERC-E 43,213 42,558 51,175 51, % 21.53% 15.00% SERC-N 42,540 40,751 48,910 51, % 25.80% 15.00% SERC-SE 47,762 45,534 60,062 60, % 33.08% 15.00% SPP 51,936 51,184 65,083 65, % 27.00% 12.00% Texas RE-ERCOT 71,416 68,548 80,510 85, % 24.07% 13.75% WECC-AB* % 43.44% 11.03% WECC-BC* % 12.39% 12.10% WECC-CAMX 54,774 53,027 63,765 63, % 20.43% 16.16% WECC-NWPP-US 50,013 48,794 62,374 62, % 28.23% 16.32% WECC-RMRG 12,392 11,847 15,364 15, % 29.68% 14.14% WECC-SRSG 23,207 22,787 29,094 29, % 27.68% 15.82% 42 Per WECC's request, data is not presented publically for Alberta and British Columbia subregions. 44

54 Chapter 4: Reliability Assessment Trends and Analysis Eastern 612, , , , % 27.05% - Interconnection Québec 38,150 35,982 41,217 42, % 17.61% 12.20% Interconnection ERCOT 71,416 68,548 80,510 85, % 24.07% 13.75% Interconnection Western 155, , , , % 26.44% 15.37% Interconnection TOTAL-NERC 876, ,713 1,040,199 1,063, % 26.29% - *Winter Peaking System Table 4.2: Peak Season 2021(S) / (W): Projected Demand, Resources, & Planning Reserve Margins Assessment Area / Demand (MW) Resources (MW) Reserve Margins (%) Reference Interconnection Total Net Anticipated Prospective Anticipated Prospective Margin Internal Internal Level FRCC 50,461 47,256 58,379 59, % 25.79% 15.00% MISO 130, , , , % 26.17% 15.20% MRO-Manitoba 4,685 4,685 6,412 5, % 24.74% 12.00% Hydro* MRO-SaskPower* 3,901 3,816 4,872 4, % 29.71% 11.00% NPCC-Maritimes* 5,622 5,350 6,661 6, % 25.89% 20.00% NPCC-New England 26,816 26,438 31,330 32, % 22.99% 15.93% NPCC-New York 33,555 32,307 40,727 43, % 34.56% 15.00% NPCC-Ontario 22,479 21,878 26,235 26, % 20.17% 17.00% NPCC-Québec* 39,415 37,097 42,746 43, % 18.19% 12.70% PJM 157, , , , % 52.54% 16.50% SERC-E 46,126 45,454 54,798 55, % 21.76% 15.00% SERC-N 43,800 42,105 50,177 52, % 24.59% 15.00% SERC-SE 49,325 47,065 62,126 62, % 33.16% 15.00% SPP 53,779 52,868 64,046 64, % 22.52% 12.00% Texas RE-ERCOT 74,966 72,098 86, , % 41.86% 13.75% WECC-AB* 13,198 13,198 16,439 19, % 50.80% 11.03% WECC-BC* 12,242 12,242 13,757 13, % 12.38% 12.10% WECC-CAMX 54,162 52,455 63,626 63, % 21.68% 16.16% WECC-NWPP-US 51,693 50,498 64,879 65, % 29.29% 16.32% WECC-RMRG 13,194 12,585 15,230 15, % 20.45% 14.14% WECC-SRSG 24,978 24,623 29,182 29, % 19.18% 15.82% Eastern 628, , , , % 32.70% - Interconnection Québec 39,415 37,097 42,746 43, % 18.19% 12.70% Interconnection ERCOT 74,966 72,098 86, , % 41.86% 13.75% Interconnection Western 160, , , , % 25.78% 15.37% Interconnection TOTAL-NERC 903, ,003 1,070,632 1,150, % 31.60% - *Winter Peaking System 45

55 Chapter 4: Reliability Assessment Trends and Analysis Table 4.3: Peak Season 2026(S) / (W): Projected Demand, Resources, & Planning Reserve Margins Assessment Area / Demand (MW) Resources (MW) Reserve Margins (%) Reference Interconnection Total Net Anticipated Prospective Anticipated Prospective Margin Internal Internal Level FRCC 52,803 49,499 60,976 62, % 27.23% 15.00% MISO 134, , , , % 18.98% 15.20% MRO-Manitoba 4,821 4,821 6,412 5, % 23.82% 12.00% Hydro* MRO-SaskPower* 4,159 4,074 5,152 5, % 30.75% 11.00% NPCC-Maritimes* 5,518 5,248 6,655 6, % 28.13% 20.00% NPCC-New England 27,218 26,841 31,353 32, % 21.23% 15.93% NPCC-New York 34,056 32,808 40,727 43, % 32.51% 15.00% NPCC-Ontario 22,265 21,056 24,646 24, % 17.96% 16.00% NPCC-Québec* 40,625 38,307 42,746 43, % 14.46% 12.70% PJM 161, , , , % 48.61% 16.50% SERC-E 49,309 48,625 58,863 59, % 22.18% 15.00% SERC-N 45,690 44,146 53,247 55, % 25.79% 15.00% SERC-SE 52,083 49,810 62,636 63, % 26.84% 15.00% SPP 56,048 55,144 62,592 63, % 14.27% 12.00% Texas RE-ERCOT 78,572 75,704 86, , % 34.91% 13.75% WECC-AB* 14,304 14,304 16,424 19, % 38.97% 11.03% WECC-BC* 13,040 13,040 14,316 15, % 17.38% 12.10% WECC-CAMX 54,005 52,298 61,639 59, % 14.53% 16.16% WECC-NWPP-US 53,294 52,101 65,079 65, % 25.84% 16.32% WECC-RMRG 14,094 13,459 16,088 16, % 19.53% 14.14% WECC-SRSG 27,424 27,069 31,763 31, % 17.95% 15.82% Eastern 650, , , , % 28.98% - Interconnection Québec 40,625 38,307 42,746 43, % 14.46% 12.70% Interconnection ERCOT 78,572 75,704 86, , % 34.91% 13.75% Interconnection Western 166, , , , % 19.92% 15.37% Interconnection TOTAL-NERC 935, ,662 1,073,175 1,152, % 27.23% - *Winter Peaking System 46

56 Chapter 4: Reliability Assessment Trends and Analysis Demand and Energy To better understand the demand requirements of an assessment area, both the seasonal on-peak demand and annual energy requirements must be studied. NERC analyses a ten-year compounded annual growth rate (CAGR) 43 for both demand and energy using forecasted data from 1990 to the present. Both Figure 4.1 and Figure 4.2 below show a consistent downward trend in demand and energy forecast data. The 2016 LTRA reference case calculates a compounded annual growth rate of 0.73 percent for summer demand, 0.72 percent for winter demand, and 0.71 percent for annual net energy. Figure 4.1: Compound Annual Growth Rate Demand *Prior to the 2011LTRA, the initial year of the 10-year assessment period is the report year (e.g., the 10-year assessment period for the 1990LTRA was ). The 2011 LTRA and subsequent LTRAs examine the initial year of the assessment period as one year out (e.g., the 10-year assessment period for the 2011 LTRA is ). The decreasing growth rate in energy is most notably detected in 2010; this is a deviation from the nearly flat energy growth rate observed in prior years. In the last year alone, the growth rate dropped by half from 0.8 to 0.7 percent. Figure 4.2: Compound Annual Growth Rate Energy Compounded annual growth rate (CAGR) provides the year-over-year growth rate over the duration of the assessment period. It is derived as follows: CAGR = (Year 10 TID / Year 1 TID)ᶺ(1 / 9) Year Growth Rate starting in 2011; only 9 years were used prior to

57 Chapter 4: Reliability Assessment Trends and Analysis Most assessment areas continue to experience a flattening growth rate in both their ten-year peak demand and energy forecasts. This is largely due to widespread implementation of energy efficiency and conservation programs, DSM, and increasing installations of distributed energy resources (DERs) that are nonobservable by utilities and treated as passive load modifiers. Figure 4.3 shows the compounded annual demand growth rate by assessment area. Figure 4.3: GR-Map: Compound Annual Growth Rate by Assessment Area Demand Historically, while utilities have worked to implement a variety of programs to reduce their peak load obligation; these reductions in energy forecasts also point to a growing change to the system as the currently metered energy needed in future years is decreasing. This is verified by examining the energy reported in past years. Figure 4.4 shows the actual energy served from 2010 to The calculated trend line shows a decrease from the 4,555 TWhs used in 2010 to 4,526 TWhs in 2015; this is a reduction of 29 TWhs or 0.64 percent. Figure 4.4: Actual Energy Used to Serve Load As a majority of renewable energy is not generated across the peak. The increasing amounts of behind-the-meter and renewable generation will continue to decrease net energy used to serve load while not similarly decreasing peak load obligations. This trend should continue unless the energy generated by 48

58 Chapter 4: Reliability Assessment Trends and Analysis behind-the-meter generation is both observable and contributing to energy profiles instead of being treated as a passive load modifier. Economic conditions also have an important impact on annual energy usage. New energy efficiency programs are still a key component for assessment areas to use to manage both peak demand and energy throughout the year. Figure 4.5 below shows the 2016 LTRA reference case projections of new programs expected to decrease overall system peak load demand by 26.3 GW through the end of the assessment period, or approximately 2.8 percent of the overall GW of expected peak load. Figure 4.5: Energy Efficiency Projections by Interconnection through 2026 Demand-Side Management There are a variety of demand response (DR) programs that contribute to an assessment area s ability to manage load; these may consist of behind-the-meter supply resources and/or load-reducing programs that are available at specific times of the year. For the purposes of this assessment, only the expected amount that are likely to respond when called to reduce peak load are collected. Each assessment area may have different mechanisms in place for accounting for this available DR and forecasting these program s availability ten years out. Any significant changes to these forecasting methods are presented and reviewed annually by the Reliability Assessment Subcommittee. Figure 4.6 below shows the total system DR considered available for the first year s summer and winter peak from each of the last six LTRAs. A decade ago, when there was a strong focus on the development of DR programs, growth projections were high for the new programs. Comparing the total, system-wide DR for year one of the 2011 LTRA and the 2016 LTRA yields a drop in the summer from 44.0 GW to 31.9 GW (27.5 percent) and in the winter from 42.7 GW to 21.6 GW (49.4 percent). Contributing factors to this decline include increased energy efficiency and solar installations. These have reduced the amount of discretionary load that is available to be reduced on demand, maturation of DR programs and their participants understanding of them, and regulatory and market rule changes that apply to DR. 49

59 Chapter 4: Reliability Assessment Trends and Analysis Figure 4.6: Demand Response Available in Year One of 2011LTRA 2016 LTRA In October of 2008, FERC issued Order No. 719, 45 which required wholesale markets to accept most DR bids in all markets. In March 2011, FERC issued Order No. 745, 46 which required that wholesale energy markets provide equal compensation to DR providers for conserving energy at the same market rate as generators are paid for producing it. Order No. 745 was brought to the Court of Appeals for the District of Columbia Circuit and challenged as beyond the authority of FERC. Although the decision of the Court of Appeals and subsequent appeal to the Supreme Court created some uncertainty for the future of market-based DR programs, available DR does not appear to have been significantly affected. In January of 2016, the Supreme Court ruled that the Federal Power Act does authorize FERC to regulate the sale of electric energy at wholesale in interstate commerce and to ensure that rules or practices affecting wholesale rates are just and reasonable. 47 The moderate changes shown in the annual forecasts of enrollments of DR in the past three years may reflect changes to market rules or market structures, improvements in technologies to facilitate participation, and local utility commission drivers to increase load management capabilities through DR. Generation Fuel Mix Examining the existing and projected generation mix is crucial to assessing potential risks to reliability. Specifically, whether or not the projected capacity additions will provide adequate levels of essential reliability services (ERS) components to support the overall state of the system. A total of 1,280 GW of nameplate capacity are expected to be available to serve load by the end of Figure 4.7 shows this system-wide, 1,280 GW of anticipated nameplate capacity by generation type. Many additional generating units are planned for the next ten years to meet a combination Figure 4.7: 2016 Existing & Tier 1 Nameplate Capacity by Fuel Type 45 FERC Order No. 719: Wholesale Competition in Regions with Organized Electric Markets 46 FERC Order No. 745: Demand Response Compensation in Organized Wholesale Energy Markets 47 Supreme Court Decision: January 25,

60 Chapter 4: Reliability Assessment Trends and Analysis of demand growth and the need for replacement generation as other units retire. Figure 4.8 shows the system s aggregated planned additions separated by generation type and planned tier designator. The three tiers, shown in Figure 4.8, indicate some degree of certainty for each unit. Generation in Tier 1 is considered to be very certain and can be expected to be available for the assessment while Tier 2 is less certain and Tier 3 is not included in a reserve margin analysis. While there are many resource types in the queue for resource planning expectations, a significant majority of all planned generation will rely on natural gas. Figure 4.8: Planned Nameplate Capacity Additions by Generation Type and Tier Table 4.4, Table 4.5, Table 4.6, and Table 4.7 show the total nameplate capacity for each fuel type by Interconnection. These include actual values reported for 2015 and all planned additions and confirmed retirements projected through Table 4.4: Eastern Interconnection Total Anticipated Nameplate Capacity by Fuel Type 2015 Nameplate (MW) 2026 Nameplate (MW) Capacity Change (MW) Capacity Change (%) Biomass 6,379 6, % Coal 242, ,825 (8,546) -3.5% Geothermal Hydro 43,736 44,845 1, % Natural Gas 332, ,298 53, % Nuclear 104, ,238 1, % Other % Petroleum 50,950 48,814 (2,136) -4.2% Pumped Storage 16,165 16, % Solar 1,634 6,233 4, % Wind 46,324 57,424 11, % Total 846, ,563 61, % 51

61 Chapter 4: Reliability Assessment Trends and Analysis Table 4.5: Texas Interconnection Total Anticipated Nameplate Capacity by Fuel Type 2015 Nameplate (MW) 2026 Nameplate (MW) Capacity Change (MW) Capacity Change (%) Biomass % Coal 20,796 20, % Hydro % Natural Gas 50,114 58,110 7, % Nuclear 5,268 5, % Solar 288 2,053 1, % Wind 15,909 26,934 11, % Total 93, ,915 20, % Table 4.6: Québec Interconnection Total Anticipated Nameplate Capacity by Fuel Type 2015 Nameplate (MW) 2026 Nameplate (MW) Capacity Change (MW) Capacity Change (%) Biomass % Hydro 40,943 41, % Natural Gas % Petroleum % Wind 3,260 3, % Total 45,536 47,048 1, % Table 4.7: Western Interconnection Total Anticipated Nameplate Capacity by Fuel Type 2015 Nameplate (MW) 2026 Nameplate (MW) Capacity Change (MW) Capacity Change (%) Biomass 3,616 3, % Coal 38,379 36,245 (2,134) -5.6% Geothermal 3,862 4, % Hydro 68,736 69, % Natural Gas 101, ,872 3, % Nuclear 7,679 7, % Other 2,962 2, % Petroleum 1,143 1,133 (10) -0.9% Pumped Storage 2,450 3, % Solar 9,476 16,907 7, % Wind 21,118 22,715 1, % Total 261, ,101 12, % 52

62 Chapter 4: Reliability Assessment Trends and Analysis Transmission Adequacy Maintaining sufficient transmission capacity is a key component of understanding and analyzing an assessment area s transmission adequacy. Load and resources are subject to a variety of factors that could lead to rapid changes to electric transmission infrastructure. This is generally restricted by slow planning, siting, and construction. While many generating units do require years to plan and build, unexpected retirements and the addition of generation with much shorter build times can stress the current transmission system. Through modeling and power flow studies, system planners provide the foundation for these essential transmission projects to be developed. A FERC technical conference was held in August of 2016 that discussed competitive transmission development processes wherein Panel Four of this discussion involved Interregional Transmission Coordination Issues. 48 Amidst the discussion was an overview of several reports from The Brattle Group that highlighted studied transmission planning needs, trends, and recommendations. 49 As unprecedented shifts in the makeup of available generating resources and load occur, policy makers and regulators should advocate for developed processes that allow for transmission solutions that meet both reliability requirements and anticipated changes to due to environmental regulations. Tabulated below are the summarized major transmission project expansions provided in this report. FRCC The FRCC Region has not identified any major projects that are needed to maintain or enhance reliability during the planning horizon. Planned projects, shown in Table 4.8, are primarily related to expansion in order to serve forecasted growing demand, and they are related to maintaining the reliability of the BES in the longer-term planning horizon or for resource integration. Table 4.8: FRCC Planned Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Levee Midway Florida Power & Light Company Reliability kV (ac) 2023 MISO MISO s Transmission Expansion Plan 50 (MTEP15) includes proposals for over $2.75 billion 51 in transmission infrastructure investment through 2024, and these fall into the following categories: 90 Baseline Reliability Projects (BRP) totaling $1.2 billion: BRPs are required to meet NERC reliability standards. 12 Generator Interconnection Projects (GIP) totaling $73.6 million: GIPs are required to reliably connect new generation to the transmission grid. 1 Market Efficiency Project (MEP) totaling $67.4 million: MEPs meet Attachment FF requirements for reduction in market congestion. 242 Other Projects totaling $1.38 billion: Other projects include a wide range of projects, such as those that support lower-voltage transmission systems or provide local economic benefit but do not meet the threshold to qualify as Market Efficiency Projects. 48 FERC Docket No. AD ; Notice Inviting Post-Technical Conference Comments; August 3, The Brattle Group: Well-Planned Electric Transmission Saves Customer Costs: Improved Transmission Planning is Key to the Transition to a Carbon-Constrained Future; June 6, MISO's Transmission Expansion Plan 51 The MTEP15 report and project totals reflect all project approvals during the MTEP15 cycle, including those approved on an out-of-cycle basis prior to December

63 Chapter 4: Reliability Assessment Trends and Analysis Several of MISO s major transmission projects are shown in Table 4.9. Table 4.9: MISO Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Great Northern Transmission Line partial segment Minnesota Power (Allete, Inc.) Hydro Integration kV (ac) 2020 MVP Portfolio 1 Ellendale to Big Stone South MVP Portfolio 1: N LaCrosse N Madison-Cardinal- Eden-Hickory Creek Great Northern Transmission Line partial segment- MVP Portfolio 1: Lakefield Jct. Winnebago Winco Kossuth County & Obrien County Kossuth County Webster Otter Tail Power Company American Transmission Co. LLC Minnesota Power (Allete, Inc.) Ameren Services Company Reliability kV (ac) 2019 Reliability kV (ac) 2024 Hydro Integration kV (ac) 2020 Reliability kV (ac) 2018 Manitoba Hydro Manitoba Hydro has plans for a significant number of system enhancement projects, including those listed in Table Manitoba Hydro is planning for an addition of the third 2,000 MW Bipolar HVdc transmission system in Bipole III provides an alternative path to serve Manitoba load in the event of a major station loss or corridor loss associated with Bipole I and II. Manitoba Hydro is expecting a new 500 kv interconnection from Dorsey to Iron Range (Duluth, Minnesota) to come into service in 2020, as a result of an 883 MW transmission service request. Manitoba Hydro is also expecting a new 230 kv interconnection from Birtle South (Manitoba) to Tantallon (Saskatchewan) station with an in-service-date of 2020, as a result of a 140 MW transmission service request. The reliability impact of the 230 kv line is not evaluated in this assessment because a construction agreement has not been finalized with the customer yet. Table 4.10: Manitoba Hydro Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Bipole 3 Riel Manitoba Reliability kV (dc) 2018 Hydro Great Northern Transmission Line (Canadian Portion) Manitoba Hydro Reliability kV (ac) 2020 SaskPower Saskatchewan has several major transmission projects for reliability during near-term of the assessment period. These projects, identified in Table 4.11, are heavily dependent on load growth, and involve the construction of approximately 570 miles (918 km) of transmission lines. 54

64 Chapter 4: Reliability Assessment Trends and Analysis Table 4.11: SaskPower Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Pasqua Swift SaskPower Reliability kV (ac) 2019 Current Area Reinforcement Pasqua Swift SaskPower Reliability kV (ac) 2019 Current Area Reinforcement Aberdeen SaskPower Reliability kV (ac) 2017 Wolverine Area Reinforcement Tantallon Area SaskPower Reliability kV (ac) 2017 Reinforcement Regina Moose Jaw SaskPower Reliability kV (ac) 2020 Area Reinforcement Regina Moose Jaw SaskPower Reliability kV (ac) 2019 Area Reinforcement Aberdeen SaskPower Reliability kV (ac) 2017 Wolverine Area Reinforcement Tantallon AMBirtle line SaskPower Reliability kV (ac) 2020 Maritimes Transmission development in the Maritimes area during the assessment period includes projects shown in Table Additional projects include the installation of a 345 kv breaker in series with an existing breaker at NB s Point Lepreau terminal in Spring This was done to mitigate contingencies and reduce import restrictions from New England. During the winter of 2016/17, the installation of two undersea 138 kv cable connections, each with a capacity of 200 MVA and a length of nine miles, will be completed and will increase capacity. This was done to improve the ability to withstand transmission contingencies in the area between NB and PEI. A 475 MW High Voltage Direct Current (HVdc) undersea cable link (Maritime Link) between Newfoundland, Labrador, and NS will be installed by early This cable, in conjunction with the construction of the Muskrat Falls hydro development in Labrador, is expected to facilitate the unconfirmed retirement of a 153 MW coal-fired unit in NS by mid The Maritime Link could also potentially provide a source for imports from NS into NB that would reduce transmission loading in the southeastern NB area. In addition, during the fall of 2018, a second 345/138 kv transformer will be added in parallel with an existing transformer at the Keswick terminal in NB. This is to mitigate the effects of transformer contingencies at the terminal. 55

65 Chapter 4: Reliability Assessment Trends and Analysis Table 4.12: Maritimes Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year NS-NL Tie (Newfoundla Nova Scotia) Nova Scotia Power Variable/ Renewable Integration kV (dc) 2017 Y-104 (West Royalty Church Road) Harbour East (Dartmouth East Eastern Passage) Maritimes Electric Nova Scotia Power Variable/ Renewable Integration kV (ac) 2017 Reliability kV (ac) 2018 New England Several major transmission projects for New England are identified in Table One significant 345 kv project that is important to the continuation or enhancement of system or subarea reliability is projected to come on-line during the assessment period. This project is the result of progress made by the ISO and regional stakeholders in analyzing the transmission system in New England, and then developing and implementing solutions to address existing and projected transmission system needs. The major project under development in New England is the Greater Boston project. The Greater Boston upgrades, which are certified to be in service by 2019, are critical to improving the ability to move power into the Greater Boston area, and also to move power from northern New England to southern New England. This set of upgrades includes a +/- 200 MVAR 345 kv interconnected static synchronous compensators (STATCOMs) in Maine that will also help to address concerns with the potential for system separation due to significant contingencies in southern New England. Table 4.13: New England Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Northern Pass Transmission Project Northern Pass Transmission LLC Other kV (dc) 2019 Northern Pass Transmission Project Northern Pass Transmission Project Northern Pass Transmission LLC Northern Pass Transmission LLC Other 60 (Under Ground) 320kV (dc) 2019 Other kV (ac) 2019 New York The Transmission Owner Transmission Solutions (TOTS) consists of three transmission projects in central New York, downstate New York, and New York City. The TOTS are part of the Con Edison and the New York Power Authority (NYPA) filing in response to a November 2012 Order from the NYSPSC that recognized significant reliability needs would occur if the Indian Point Energy Center (IPEC) were to become unavailable. 52 The TOTS transmission projects are described in the following three projects: Project One: The Ramapo-Rock Tavern project will establish a second 345 kv line from Con Edison s Ramapo 345 kv substation to Central Hudson Gas and Electric Corporation s (CHGE) Rock Tavern 345 kv substation. The project will increase the import capability into Southeastern New York (SENY); this includes New York City, during normal and emergency conditions and will provide a partial solution for system reliability should the IPEC retire. The 52 New York Public Utilities Commission; Order Instituting Proceeding and Soliciting Indian Point Contingency Plan; November

66 Chapter 4: Reliability Assessment Trends and Analysis project will be located in Orange and Rockland Counties in New York along the right-of-way for the existing Con Edison 345 kv Feeder 77 (Ramapo to Rock Tavern) and use existing transmission towers. The transmission line terminals are located in NYBA s Zone G. This project involves work that will be performed by Orange & Rockland Utilities (O&R) and CHGE, as such, Con Edison has and will continue to coordinate this effort with both O&R and CHGE. Project Two: The Marcy South Series Compensation project is a transmission improvement project that adds switchable series compensation to increase power transfer. This is done by reducing series impedance over the existing 345 kv Marcy South lines. Specifically, the project adds 40 percent compensation to the Marcy-Coopers Corners 345 kv line, 25 percent compensation to the Edic-Fraser 345 kv line, and 25 percent compensation to the Fraser-Coopers Corners 345 kv line through installation of series capacitors. The project also involves upgrades at Marcy and Fraser 345 kv substations. These upgrades involve reconductoring approximately 21.8 miles of the NYSEG-owned Fraser-Coopers Corners 345 kv line (FCC-33) with a higher thermal-rated conductor. The project increases thermal transfer limits across the Total East Interface and the UPNY/SENY Interface. Project Three: The third project splits an existing feeder between Goethals and Linden Cogen substations, and it will provide a similar solution at a lower cost and with lower environmental impacts. This project is located in Staten Island and Brooklyn, New York; and Union County (Linden), New Jersey. Ontario Several major transmission projects for Ontario are identified in Table Northwestern Ontario is connected to the rest of the province by the 230 kv double-circuit East West Tie. Local load growth is forecasted as a result of an active mining sector in the region. To address load growth, additional capacity is required to maintain reliable supply to this area under the wide range of possible system conditions. Anticipated to be in service in 2020, the expansion of the East West Tie with the addition of a 230 kv new double-circuit transmission line will provide reliable and cost-effective long-term supply to the Northwest. Table 4.14: Ontario Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year QFW Hydro One Economics/ Congestion kV (dc) Delayed/ Unknown East-West Tie Hydro One Reliability kV (dc) 2020 West of Thunder Bay Lines Hydro One Reliability kV (ac) 2022 Forecasted demand growth in the areas west of Thunder Bay and north of Dryden will require increased transfer limits west of Mackenzie Transformer Station (TS). Development work is proceeding for a new 230 kv doublecircuit line between Lakehead TS and Mackenzie TS, and a new single-circuit line between Mackenzie TS and Dryden TS. Forecasted demand growth in the Ottawa area will require reinforcements to the transmission system to relieve future thermal constraints. Plans including line reconductoring are already underway to address the thermal constraints. Ontario is monitoring the progress of the continued operation of nuclear units at Pickering Nuclear Generating Station. Pickering Nuclear Generating Station units connect directly to the 230 kv system at Cherrywood Transformer Station, which is located in the east side of the greater Toronto area. The retirement of Pickering Nuclear Generating Station requires an additional 230 kv supply source for the area. This will be provided by the 57

67 Chapter 4: Reliability Assessment Trends and Analysis new Clarington 500/230 kv transformer station with a planned 2018 in-service date. Clarington Transformer Station will also improve load restoration capabilities east of Cherrywood following certain contingencies. Bulk power transfers into the GTA from the west are expected to increase as a result of the planned shutdown of Pickering Generating Station, major refurbishment of other nuclear generating units, and the incorporation of significant amounts of renewable generation in Ontario. Because of the increased bulk power transfers and increasing local demand, the capacity of the transmission lines between Trafalgar TS and Richview TS and the 500/230 kv transformers at Claireville TS and Trafalgar TS are forecasted to be exceeded by Planning studies are being finalized. Planning options have been assessed and are expected to include the installation of 500/230 kv autotransformers at the existing Milton Switching Station, with eight 230 kv circuit terminations, and 12 km of new double-circuit line sections connecting the new Milton TS to Hurontario Switching Station. Hydro Québec The major transmission projects in the Hydro Québec footprint are: the Romaine River Hydro Complex Integration, the Chamouchouane Montréal 735-kV Line, the Northern Pass Transmission Project, and the Champlain-Hudson Power Express Project. The Romaine River Hydro Complex Integration Construction the Romaine River Hydro Complex project is presently underway. Its total capacity will be 1,550 MW. Romaine-2 (640 MW) was commissioned in December 2014, and Romaine-1 (270 MW) in December Romaine-3 (395 MW) and Romaine-4 (245 MW) will be integrated in at Montagnais 735/315-kV substation. The Québec area is reiterating its commitment to sustainable development by focusing on renewable energy at the Romaine complex, which will help meet current needs without jeopardizing the energy supply of future generations. Main system upgrades for this project has required construction of a new 735-kV switching station Aux Outardes, which is located between existing Micoua and Manicouagan substations. Two 735-kV lines have been redirected into the new station, and one new 735-kV line (5 km or 3 miles) has been built between Aux Outardes and Micoua substations. This upgrade was commissioned in Summer The Chamouchouane Montréal 735-kV Line Planning studies have shown the need to reinforce the transmission system with a new 735-kV line in the near future in order to meet the reliability standards. The line will extend from the Chamouchouane substation on the eastern James Bay subsystem to a new substation (Judith Jasmin) in Montréal (about 400 km or 250 miles). The new 735kV substation is required to fulfill two objectives: providing a new source of electricity supply on the north shore of Montreal and connecting the new 735kV line from Chamouchouane to the Montreal metropolitan loop. This project will reduce transfers on other parallel lines on the Southern 735-kV Interface, thus optimizing operation flexibility and reducing losses. The line is scheduled for the winter peak period. Public information meetings have been held and the construction phase has begun. The Northern Pass Transmission Project This project to increase transfer capability between Québec and New England by 1,090 MW is currently under study. It involves the construction of a ±320-kV dc transmission line about 49 miles (79 km) long from Des Cantons 735/230-kV substation to the Canada United States border. This line will be extended into the United States to a new substation built in Franklin, New Hampshire. The project in Québec also includes the construction of an HVdc converter at Des Cantons and a 320-kV dc switchyard. The planned in-service date is The Champlain-Hudson Power Express Project This project to increase transfer capability between Québec and New York by 1,000 MW is currently under study. It involves the construction of a ±320-kV dc underground transmission line about 50 km (31 miles) long from Hertel 58

68 Chapter 4: Reliability Assessment Trends and Analysis 735/315-kV substation just south of Montréal to the Canada United States border. This line will be extended underground and underwater (Lake Champlain and the Hudson River) to Astoria station in New York City. The project in Québec also includes the construction of an HVdc converter at Hertel. The planned in-service date is currently under review. Upcoming Regional Projects Other regional substation and/or line projects are in the planning/permitting stages. There are about a dozen regional transmission projects in the Montréal and Québec City areas. There are another dozen in other areas with in service dates from 2016 to 2020, consisting mostly of 315/25-kV and 230/25-kV distribution substations to replace 120-kV and 69-kV infrastructures. Two of these more notable regional transmission projects are shown in Table Table 4.15: Québec Regional Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year La Romaine 3-4 Hydro-Québec Hydro kV (ac) 2017 Integration Line CHM-MTL Hydro-Québec Reliability kV (ac) 2018 SERC The major transmission projects in the SERC footprint are as follows: Construction of the Union-Tupelo and Selmer-West Adamsville 161 kv lines support voltage and the changing flows in the area. Pin Hook needs an additional 500/161 kv transformer to alleviate overloads in the SERC-N area. Construction of the Plateau 500 kv Substation will alleviate decreasing voltages and higher flows on lines caused by increased loads in the area. A new static VAR compensator (SVC) installation at the Davidson 500 kv substation will increase dynamic reactive reserves for support. Construction of the 55 mile Vogtle to Thompson 500 kv line will support the addition of future generation. New 230 kv transmission projects are under construction in conjunction with VC Summer Nuclear Units 2 and 3. SPP The SPP assessment area s 2016 Board-of-Directors-approved SPP Transmission Expansion Plan Report (STEP) provides details for proposed transmission projects needed to maintain reliability while also providing economic benefit to the end users. The 2016 SPP Transmission Expansion Plan (STEP) contains a comprehensive listing of all transmission projects in SPP for the 20-year planning horizon. These projects consist of $6.1 billion in new transmission and upgrades. Several of these major transmission projects are shown in Table Table 4.16: SPP Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Sibley Mullin TSMO Economics/ kV (ac) 2016 Creek Congestion Cherry Co. NPPD Reliability kV (ac) 2018 Gentleman Cherry Co. Holt NPPD Reliability kV (ac) 2018 Co. Tuco Yoakum SPS Reliability kV (ac)

69 Chapter 4: Reliability Assessment Trends and Analysis The Integrated Transmission Planning (ITP) process is Southwest Power Pool s iterative three-year study process that includes 20-Year, 10-Year, and Near Term Assessments. The 20-Year Assessment (ITP20), performed once every three (3) years, identifies transmission projects, generally above 300 kv, needed to develop a grid flexible enough to provide benefits to the region across multiple scenarios. The 10-Year Assessment (ITP10), performed once every three (3) years, focuses on facilities 100 kv and above to meet system needs over a 10-year horizon. The Near-Term Assessment (ITPNT), performed annually, assesses system upgrades, at all applicable voltage levels, required in the near-term planning horizon to address reliability needs. The goal of transmission integration studies 53 to evaluate the adequacy of each Integrating Entity s transmission facilities at the time of their integration date and whether they are in compliance with NERC Reliability Standards, SPP Criteria, and Transmission Owner-specific planning criteria (if a waiver allows for the SPP or NERC Criteria to be superseded). An initial integration study was conducted in 2013 for the Integrating Entities that identified projects needed before and after the October 2015 integration date. This latest study was a refresh of the initial study to determine if any supplemental issues emerged when considering more current information. SPP leveraged the 2016 ITP Near-Term model set as the starting point for the analysis. The Integrating Entities provided updates to topology, generation dispatch, and load information. Texas RE-ERCOT Several of Texas RE-ERCOT s major transmission projects are shown in Table The recently updated ERCOT future transmission projects list includes the additions or upgrades of 3,954 miles of 138-kV and 345-kV transmission circuits, 24,159 MVA of 345/138-kV autotransformer capacity, and 3,005 MVar of reactive capability projects. These are planned in the TRE-ERCOT Region between 2016 and Table 4.17: TRE-ERCOT Major Transmission Projects Name Company Driver Line Length (Circuit Miles) Operating Voltage/Type Expected In- Service Year Lobo to North ETT Reliability 345kV (ac) 2016 Edinburg: Construct 345 kv Line Add second circuit SHRY Economics/ 345kV (ac) 2018 to SLU panhandle loop Congestion New 345kV line from Loma Alta to N Edinburgh SLU Reliability 345kV (ac) 2016 A new Houston Import Project, 130-mile 345 kv double circuit line (each circuit rated at 5000 Amps) from Limestone to Gibbons Creek to Zenith, is planned to be in service before the summer peak of The Houston area is one of the two largest demand centers in the ERCOT system and the fourth largest city in the United States. The Houston area demand is met by generation located within the area and by importing power via high-voltage lines into the area from the rest of the ERCOT System. This new line will support anticipated long-term load growth 53 SPP Integrated Transmission Planning 54 ERCOT: Houston Import RPG Project; April 8,

70 Chapter 4: Reliability Assessment Trends and Analysis in the Houston region. Power imports into the Houston area are expected to be constrained until the new import line is constructed. In July 2014, the owners of the Frontera generation plant, a 524 MW natural gas facility located on the west side of the Lower Rio Grande Valley (LRGV), announced that they were planning to switch part of the facility (170 MW) out of the ERCOT market in This entire facility would no longer be available to ERCOT in In June, 2016, the ERCOT Board of Directors endorsed the reliability need for the two 300 MVAr SVCs located inside the LRGV to be in service prior to summer of 2021 to meet ERCOT and NERC reliability criteria for the LRGV. 61

71 Chapter 5: Additional Reliability Issues NERC continues to monitor and report on a variety of other issues that are generally categorized as lower risk. While these may not require immediate attention or action, there is a consistent need to assess all system changes or impacts to be aware of any risks before they develop. The 2016 LTRA identifies the following items as issues, trends, and events that warrant further attention and study: EPA Clean Power Plan (CPP) Grid energy storage System short circuit strength Modeling Reactive power requirements for nonsynchronous generation System restoration Reactive power supporting devices 2017 solar eclipse Clean Power Plan On August 3, 2015, the EPA issued its final rule, Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Generating Units. 55 Initial compliance of the final rule was set to begin in 2022 with final compliance in The final rule aims to cut CO₂ emissions from existing power plants to 32 percent below 2005 levels by NERC conducted an analysis of the final rule in order to assess potential reliability risks to the BPS as a result of the rule. 56 As a result of the analysis, NERC determined that already occurring changes to the resource mix would accelerate if the final rule were to be implemented. Among NERC s key findings of its analysis are the following: The CPP is expected to accelerate a fundamental change in the electricity generation mix in the United States and transform grid-level reliability services, diversity, and flexibility. Integration of large amounts of renewables are expected to occur on the BPS regardless of the CPP. The CPP is expected to further flatten annual energy demand growth. Resource mix changes have regional significance, spurring the need for additional transmission and pipeline infrastructure. NERC recommended that, due to the wide ranging effects of the CPP, planning processes should already be underway to ensure that requisite transmission and pipeline infrastructure be built in a timely manner. NERC further recommended that planning coordinators and transmission planners should conduct system reliability evaluations to identify areas of concern. NERC continues to hold itself out as a resource for states and planners as state submittals are being formulated. Finally, NERC also recommended that work should be continued around sufficiency guidelines for essential reliability services (ERSs) and evaluation of the effects that distributed energy has on the BPS. On February 9, 2016, the Supreme Court stayed implementation of the CPP pending judicial review. The stay will remain in effect through the review of the CPP by the Court of Appeals for the District of Columbia Circuit (D.C. 55 EPA: Clean Power Plan for Existing Units 56 NERC Potential Reliability Impacts of EPA's Proposed Clean Power Plan 62

72 Chapter 5: Additional Reliability Issues Circuit) and until the Supreme Court decides the matter in the event that it is ultimately appealed to the Supreme Court. This legal process could continue into the middle of The stay of the final rule has an impact on the ultimate validity of the final rule as well as the timing of it should it ultimately be upheld by the courts. It is also important to note that NERC determined that many of the changes occurring on the BPS are occurring regardless of the CPP and should therefore be incorporated into system and operational planning. System Short-Circuit Strength The safe, reliable operation of electrical power systems requires the ability to predict and model the sources of fault current in order to select equipment properly rated for the required duty and to properly set protective relays for selective operation. Nonsynchronous powered generating plants are also sources of fault current and are considered in addition to typical synchronous generating sources. Synchronous machines in the electric power grid, their operation, and respective short-circuit behavior have been established and are well understood in comparison to some types of nonsynchronous power plant fault current and performance. A nonsynchronous power plant is separated from conventional generation by its unique shortcircuit behavior. The unique short-circuit characteristic for nonsynchronous machines emanates from either when an induction generator is directly connected to the grid or when the nonsynchronous machine is decoupled from the grid through power electronic devices (e.g., inverters). Therefore, accurate short-circuit studies are needed to determine that the maximum short-circuit contribution from a given machine is within the limits of the circuit breakers and that protective devices are coordinated to function properly over a specific range of potential conditions. Short-Circuit Fault Contribution of Nonsynchronous Resources The short-circuit fault contribution from large nonsynchronous plants that are connected to the transmission voltages are a primary concern to safe and reliable operations of the bulk electric power system. Nonsynchronous plants of interest typically consist of multiple wind turbines/photovoltaic (PV) systems connected to transmission facilities (greater than 100kV). Here, the importance of both balanced and unbalanced short-circuit fault analysis to determine the worst case fault given select components of a nonsynchronous plant is discussed. This section does not focus on nonsynchronous collector systems nor internal plant protective relaying problems. Nonsynchronous Plant Types A significant difference between a nonsynchronous plant and a typical power plant is the total number of machines/units employed at a transmission bus. A typical conventional power plant (i.e., combustion, hydro, or steam) might consist of a single unit or a few large units. Therefore, the components for short-circuit modeling of a conventional power plant includes each generator and its respectively sized step-up transformer. In contrast, a nonsynchronous power plant of similar MW size to the conventional generator will consist of many machines/units, their individual step-up transformers, a medium voltage collector system (i.e., cabling), and a substation transformer in order to be modeled correctly for short circuit studies. Industry has divided large megawatt rated wind turbines into five different groups based on their machine type, speed control capabilities, and operational characteristics. The following list provides the wind turbine groups by their type and associated machine: Type I, squirrel cage induction generator Type II, squirrel cage wound rotor induction generator with external rotor resistance Type III, double fed asynchronous generator Type IV, full power converter generator (PV/wind) Type V, synchronous generator mechanically connected through a torque converter 63

73 Chapter 5: Additional Reliability Issues For short-circuit fault current, Types I through IV are of greater concern than Type V wind turbines. The detailed behavior and characteristics of fault contributions of nonsynchronous plants can be very specific to a particular turbine design. Turbine manufacturers must provide accurate information on balanced and unbalanced fault performance for the particular turbines in a specific plant. Photovoltaic (PV) systems generally use a full power dc-ac converter and typically produce similar fault characteristics to the Type IV wind turbine. System Strength As the number of inverter-based resources increases, and as the number of dispatched synchronous sources decreases, there will be an operating point when the grid is no longer strong enough to support stable operation of the power electronic converters connected within the wind and PV plants. Very few systems have faced this issue in actual operation (e.g., South Texas Sub-Synchronous Resonance Event of 2009). Knowledge of this issue is built upon converter performance tests and detailed analysis using transient simulation tools, such as Power Systems Computer Aided Design (PSCAD) and Electromagnetic Transients Program (EMTP). Since such tools and analytical methods are not well suited to studying large-scale risks for many plants over wide geographic areas, the challenge is to take what is learned from detailed analysis of a few plants and extend that learning across larger regions using more practical methods. Short circuit ratio (SCR) is a calculation used to screen for weak grid conditions near power electronic converters. This method has been borrowed from screening for weak grid conditions near HVdc converters and is currently being applied to nonsynchronous plants. 57 An adequate SCR for today's nonsynchronous inverter designs is defined at the point of interconnection and typically has a calculated ratio in the range of 3-5, where 3 is the minimum ratio to be considered sufficiently robust and 5 is considered vigorous. When the calculated SCR at a plant interconnection point is lower than 3, there is significant concern that the internal plant controls will not function in a stable manner (i.e., the positive sequence stability representation of the plant may not represent the true behavior of the plant or be mathematically stable). Low SCRs increases the chance of subsynchronous behavior and control interactions among neighboring devices employing power electronics. 58 The low SCR problem is typically identified and addressed during interconnection study stages of a nonsynchronous plant. This issue is fundamentally a local problem and can be remedied with upgrades, such as synchronous condensers. However, it is an ongoing issue because as synchronous generators retire or network topology changes take place, it is possible that an area in which nonsynchronous generators have been interconnected can evolve into a weaker system. As a result, the SCR should be used to re-evaluate the strength of the interconnection points due to temporal network modifications. The SCR ratio is not included in daily system operations as it is not achievable to determine critically diminished points in the system and then be able to implement immediate corrective actions (i.e., install equipment). This is to ensure a sufficiently robust solution over a reasonable range of typical operating conditions. Modeling NERC is committed to assessing the quality of the power system models used to plan the BPS. NERC is also committed to support the development and advancement of models and modeling practices to ensure that longterm and short-term planning engineers have the tools and capabilities to plan and operate the BPS. Currently under evaluation or a development plan are case quality metrics, dynamic load modeling, and adequate modeling of DERs. 57 SCR is the ratio of the available system strength (measured in short circuit MVA) to the MW rating of the wind or PV plant. 58 ERCOT: System Strength Assessment of the Panhandle System

74 Chapter 5: Additional Reliability Issues Case Quality Metrics NERC performed its Phase 2 Assessment of Case Quality Metrics 59 future-year planning cases for the Eastern, Western, and Texas Interconnections. This included reassessment of the Phase 1 metrics 60 as well as development of new metrics for Phase 2, particularly with respect to dynamic models. NERC is working with the case creation entities designated, pursuant to MOD-032, to incorporate the metrics (as appropriate) for their interconnections to improve the performance scores for their models moving forward. A number of dynamics case modeling issues were identified in the Phase 2 assessment, such as power development fractions for turbine-governor models, generator time constants and inertia constants, and generator saturation factors. Reassessment of the Phase 1 metrics showed no adverse trends between the two case years, and NERC will continue analyzing case quality to build these trends over the longer-term. Dynamic Load Modeling Dynamic load models, such as composite load models, are capable of capturing the dynamic response of various end-use loads, namely induction motor load as required per TPL The NERC Load Modeling Task Force 62 (LMTF) is supporting the development and robust implementation of these models as well as the phased adoption of these models while gaining experience with the model in stability studies. LMTF has created a forum for utility planning engineers to share experiences and modeling efforts with other utility engineers, software developers, and subject matter experts. End-use loads are rapidly changing due to energy efficiency standards and economics. Grid friendly loads that exhibit electrical characteristics that support the power grid during abnormal conditions (such as faults) are being replaced with electronically coupled loads controlled by converter technology. These electronically coupled loads may not exhibit this grid friendly characteristic; rather, they tend to have controls that maintain constant power consumption regardless of system voltage or frequency (with current limiters for protection purposes). The makeup and characteristics of end-use load technology are continually and rapidly evolving with the continued penetration of electronically coupled loads such as electric vehicles, plug-in electric hybrids, higher efficiency single-phase air conditioners, compact fluorescent lighting, LED lighting, LCD and LED televisions, variablefrequency drives, and electronically commutated motors. NERC is also coordinating with the electric utility industry to understand the end-use load response needed for future reliability of the electric grid such that the BPS maintains stable equilibrium for major grid events. Preliminary studies have developed approaches for the ideal response of large-power electronic (electronically coupled) loads, such as electric vehicle chargers. In addition, NERC has been a contributor to the development of IEEE 1547, 63 particularly with respect to sharing BPS reliability perspectives and the impact that aggregated DERs can have on BPS performance. Distributed Energy Resource Modeling As the penetration of DERs continues to increase across the North American BPS, it becomes increasingly important to ensure that steady-state and dynamic models are able to sufficiently represent the individual or aggregate response of DERs for planning and operations purposes. The NERC Essential Reliability Services Working Group (ERSWG) and LMTF technical groups are exploring the modeling practices used for capturing these resources and any modeling improvements or recommended modeling practices for the electric utility industry to consider. While the industry is still learning much from areas with high penetration of DERs, including international experience, NERC supports information sharing and proactive exploration of the tools, models, and practices to 59 NERC Phase 2 Case Quality Metrics 60 NERC Case Metrics Summer Base Case Quality Assessment; Phase I - Powerflow and Dynamics Case Quality Metrics 61 NERC Standard TPL Transmission System Planning Performance Requirements 62 NERC Load Modeling Task Force 63 IEEE 1547 Standard for Interconnecting Distributed Resources with Electric Power Systems 65

75 Chapter 5: Additional Reliability Issues help ensure reliability of the BPS moving forward. The ERSWG and LMTF are preparing technical materials developed by these industry stakeholder groups to share with the industry. Reactive Power Requirements for Nonsynchronous Generation: FERC Order 827 FERC issued Order No on June 16, 2016, eliminating the exemptions for wind generators from the requirement to provide reactive power by revising the pro forma Large Generator Interconnection Agreement (LGIA), Appendix G of the LGIA, and the pro forma Small Generator Interconnection Agreement (SGIA). While some ISOs have reactive power standards for variable energy resources (VERs), FERC Order 827 states that all new interconnecting nonsynchronous generators will be required to provide reactive power at the high-side of the generator substation as a condition of interconnection. FERC found that, due to technological advancements, the cost of providing reactive power no longer creates an obstacle to wind power development, and this decline in cost results in the current exemptions being unjust, unreasonable, and unduly discriminatory and preferential. Key items for reliability addressed by FERC in its order 65 include: Power Factor Range: All newly interconnecting nonsynchronous generators must design their Generating Facilities to maintain a composite power delivery at continuous rated power output at the high-side of the generator substation. At that point the non-synchronous generator must provide dynamic reactive power within the power factor range of 0.95 leading to 0.95 lagging, unless the transmission provider has established a different power factor range that applies to all non-synchronous generators in the transmission provider s control area on a comparable basis. Point of Measurement: Order 827 specifies the point of measurement for reactive power as the high-side of the generator substation. To clarify this location, the Commission states: As an example, the generator substation would be the substation for a wind generator that separates the low-voltage collector system from the higher voltage elements of the Interconnection Customer Interconnection Facilities that bring the generator s energy to the Point of Interconnection. Dynamic Reactive Power Capability Requirements: Order 827 states that reactive power capability can be achieved by systems using a combination of dynamic capability from the inverters plus static reactive power devices to make up for losses. This gives the Generator Owner flexibility to [u]se static reactive power devices to make up for losses that occur between the inverters and the high-side of the of the generator substation, so long as the generators maintain 0.95 leading to 0.95 lagging dynamic reactive power capability at the high-side of the generator substation. Real Power Output Threshold: All newly interconnecting nonsynchronous generation must meet the reactive power requirements at all real power output levels. FERC provided an example of a 100 MW generator required to provide 33 MVAR at 100 MW output and 3.3 MVAR at 10 MW output. This essentially is a triangle-shaped capability curve based on the amount of active power being delivered at the point of measurement. NERC is developing technical guidance to support Order 827, and will include this material as part of the Reliability Guideline on Reactive Power Planning currently being developed by the System Analysis and Modeling Subcommittee under the NERC Planning Committee. 64 Federal Energy Regulatory Commission, Order No. 872, 16 June NERC provided comments on the Notice of Proposed Rulemaking (NOPR) preceding this Final Rule. 66

76 Chapter 5: Additional Reliability Issues System Restoration Past NERC assessments have identified blackstart units as a potentially emerging issue as more conventional generation type units that have traditionally provided blackstart capability are retiring than are being built across the system. Blackstart units are generally smaller in output and are essential towards the implementation of system restoration plans that allow sections of the grid to return to service following a disturbance. As the determination of total installed capacity for a reserve margin calculation make no distinction between generation types, having fewer blackstart units in an area would have little to no impact on this deterministic resource adequacy metric. However, a large reduction in these units could significantly extend the duration of a blackout or otherwise limit a localized effort to mitigate immediate power quality issues. Additionally, as many new, variable, and decentralized resources are installed, there is increasing difficulty with providing stable communication between these generating units and control rooms. Since a majority of wind and solar units generate the maximum energy possible at any given moment, a growing amount of generation without controllable outputs and ability to respond to system needs further complicates maintaining system resiliency. With the retirement of these conventional units, transmission owners and other applicable registered entities should review their individual needs to restore the interconnection and maintaining system reliability. Practice and functionality for using renewables for system restoration and blackstart requirements are still at early stages of research and development. There are two approaches being considered for using VERs in system restoration: grid-forming and grid-following. This restoration research is currently focused on using the variable resources for grid-following. The likely outcome of the research will be to maintain the top-down restoration approach of energizing the high-voltage/100kv system using conventional generation and then using the VERs primarily to aid in island balancing and frequency regulation. There are challenges with using variable renewables for restoration; these resources are dependent on their energy inputs (i.e., sunlight, wind) being available during system restoration, and today s utility-scale, commercially available wind or solar PV resources were not specified and tested with the ability to start or run into a black system in mind. Thus, for existing wind and solar PV resources to participate in system restoration, they currently must follow and coordinate with a grid voltage and frequency that has been set by a synchronous generation resource. Viable, large-scale capability for blackstart with wind and solar PV are possible if this is a desired feature, but are several years away from commercial availability. NERC s 2012 LTRA 66 identified PJM as one area that had experienced a recent downward shift in blackstart-capable units. To review their own needs more thoroughly, PJM initiated the System Restoration Strategy Task Force (SRSTF) 67 to examine the current system restoration planning process. PJM did this to determine its viability and efficiency moving forward and to recommend any changes to any associated procurement, cost allocation, and compensation methods for system restoration. The task force recommended a number of changes to the existing rules for blackstart generation and the identification of critical load. It also developed the PJM RTO Wide Five-Year Selection Process Black Start RFP. The RTO-wide RFP is issued every five years and any unit that is interested in providing blackstart service can offer in to PJM s market. PJM would then review the existing blackstart units along with the new units offering into the RFP and optimize blackstart generation throughout the RTO. 68 Reactive Power Supporting Devices There are two components to the power supplied by conventional electric generators: real power and reactive power. Real power capacity performs the work of lighting, heating, cooling, and operating motors for a variety of uses, and can be replaced either locally or very remotely. This is a characteristic distinctive to real power since it can travel long distances via the BPS without losing effectiveness. However, the reactive power necessary to support BPS voltage, and to avoid collapse as real power flows across the BPS, has to be provided locally. 66 NERC 2012 Long-Term Reliability Assessment 67 System Restoration Strategy Task Force 68 PJM Black Start / System Restoration; presentation October 5,

77 Chapter 5: Additional Reliability Issues Reactive generators will increasingly be utilized to replace dynamic voltage support lost from conventional retirements. These include SVCs; static synchronous compensators (STATCOMs); other electronic flexible alternating current transmission system devices; and synchronous condensers, which are large motors configured to provide voltage support. Low-voltage ride-through capability and effective protection and control of the reactive devices are included in minimum BPS reliability criteria, especially as reactive generator penetration increases. Current BPS inventory of these devices as time passes and undesirable events are caused or exacerbated by their inability for low voltage ride-through are worth developing and monitoring. Advanced Capabilities of New Technology Resources Deployment of new VERs, primarily wind and solar generation, has been rapid in recent years. The amount and rate that new additions are being made continues to increase yearly. Presently available technology on these resources offer the options to provide great flexibility for operation. These resources can change output power very quickly, which is extremely helpful to stabilize the frequency during disturbances and system restoration. The vast majority of new wind generation are Types 3 and 4 machines, which have capabilities to provide frequency response control. These controls have inertia-based and governor controls that provide complementary functionalities. Inertia-based Controls Most new Type 3 and 4 wind generators have built-in capabilities to provide fast frequency response. This response is based on temporarily using the stored inertial energy in the rotating mass of the wind turbine. These are often referred to as synthetic inertia controls and they respond rapidly for a frequency drop in a 1 10 second time frame. The primary function of this fast frequency response is to provide arresting power, shown in Figure 5.1. These controls use the inertial energy from the rotating wind turbine to supply power to the electric power system. Under undisturbed operation, the mechanical power input and electrical output are balanced. During a large under-frequency event when this wind generator control is enabled, the electrical output is greater than the mechanical input during the inertia response period, extracting inertial energy out of the rotor and causing the machine speed to decrease, thereby allowing the turbine to provide a very fast injection of additional power during the arresting phase of the frequency event. After the arresting period subsides and primary frequency response (PFR) action takes over, the mechanical energy must be recovered to bring the wind turbine back to the predisturbance rotational speed (and mechanical power). The turbine again reaches equilibrium when mechanical power input equals electrical power output. Because the turbine loses some efficiency during the time when it is slowed from its optimal operating points, the energy recovery during periods of moderate wind speed will typically to be on the order of twice or more arresting energy delivered. The control is progressively more effective at higher wind speeds, and the recovery energy is supplied by the wind (with little if any energy recovery period needed) when the turbine is operating at its rated output and wind speeds are sufficient to provide additional energy. Figure 5.1: Time Frames of Frequency Response 68

78 Chapter 5: Additional Reliability Issues Governor Control Most existing and new Type 3 and 4 wind turbines, solar thermal, and new solar PV resources have the built-in control capability to provide PFR for both over frequency and under frequency events in the 5 60 second response time frame. This type of control is very similar to the governor control of synchronous thermal and hydro generation. It is often called governor control, which is unlike synchronous generation that reacts to and controls turbine speed (as a proxy for grid frequency), PFR relates the size of the resource lost to the resulting net change in system frequency. This is done during the period when stabilizing frequency is determined following the initiating system disturbance event. This is illustrated in Figure 5.2. In Figure 5.2, results of an investigation of the EI show how governor controls could impact grid frequency. In Figure 5.2, a condition with a significant amount of wind generation is subjected to a very large loss of generation event. The cases (one with just governor response, in red, and one with both governor and inertial fast frequency response, in green) show a substantial improvement in the key metric, frequency nadir, over the reference case (in blue). In this case, the wind governors are deliberately set to respond with a similar speed as incumbent thermal resources. While this setting can be increased, having some fast-responding resources compared to others can result in unintended consequences with respect to system frequency. These frequency response time frames are set based on the response times of most combined-cycle units. In order to provide the governor response for under frequency events, the wind or solar resource would have to be operating in a curtailed state to retain headroom for increasing its power output. This precurtailment is not required for wind or solar to respond to over frequency events or for the synthetic inertia response from wind turbines. Figure 5.2: Illustration of Wind Turbine Frequency Controls NREL: Eastern Frequency Response Study; May

79 Chapter 5: Additional Reliability Issues Solar Eclipse North America will experience a total solar eclipse on August 21, 2017, similar to the total eclipse that passed over continental Europe, Nordic countries, and Great Britain in The path of the 2017 solar eclipse has been predicted by NASA 70 along with the affected levels of solar gradation outward from the path. A total solar eclipse is a precisely predictable event that causes substantial effects to wide-scale solar generation within a very short amount of time. The output generated by PV/solar systems will be either diminished or drastically reduced within the window of this event. Sudden widespread diminishing of solar irradiance may heavily affect areas with large amounts of utility scale PV energy installations or behind-the-meter DERs. To further examine the potential impacts of this event, NERC will perform analysis and release a whitepaper summarizing the impact of the North American Solar Eclipse on the BPS in the first quarter of The assessment will leverage studies on past eclipse events to conduct an analysis on areas with high amounts of solar penetration along the path of the eclipse. NERC will identify any reliability concerns surrounding the BPS s ability to withstand/endure the event. As the number of VERs in the power system increases, there is a greater dependency in the power system on intermittent energy sources. As a result, there is an emerging concern on maintaining a reliable and operable system during periodic astronomical events (i.e., solar eclipses, geomagnetic storms). For example, Figure 5.3 below shows the path of the upcoming North American total solar eclipse of 2017 and the future total solar eclipse of Both eclipse routes move from the west to the east direction across North America. The map above shows that the August 21, 2017, eclipse proceeds across the U.S.A. in southerly movement with first and last total eclipse observations occurring in Oregon and South Carolina respectively. The April 8, 2024, eclipse advances northerly; the total eclipse will be first viewable in Sinaloa, Mexico, and lastly visible in Newfoundland and Labrador, Canada. Although both eclipse events occur (over by 1:30 p.m.) before historical time of day peak periods, the effect of eclipses on the BPS will become more relevant as more variable generation is installed in the system. Future detailed studies and coordination may be needed to ensure the effect of astronomical events on the behavior of wide-area BPS facilities are predictable and maintainable. Figure 5.3: Solar Map Projected Trajectory of the 2017 and 2024 Solar Eclipses 70 Total Solar Eclipse of 2017 AUG 21; NASA.Gov 70

80 Chapter 6: Regional Overview This chapter provides an overview of the results of this assessment for all Regional Entities by assessment area. Through joint efforts, NERC and the Reliability Assessment Subcommittee (RAS) built the process and analyses used for this assessment. In addition to the collection of data, NERC guides the creation of a set of narrative questions that seek to explain and clarify the potential risks to reliability that are identified throughout the BPS. Both the data and the narrative responses undergo a thorough peer review process within the RAS on an assessment area level and by the Regional Entities in a separate narrative question and review period. Presented here are highlights of the emerging issues and details by assessment area obtained through this comprehensive process. Highlights of Emerging Issues An intrinsic component of the peer-reviewed narratives is identifying emerging issues and potential risks to reliability that have been studied through additional assessments. Detailed here are the highlights of these emerging issues by assessment area. FRCC Weather events in the Gulf of Mexico could potentially have an impact on the availability and transportation of natural gas. However, dual-fuel capability, the increase of onshore (outside of Florida) gas resources, and a third gas pipeline currently under construction in central Florida (in service mid-2017) would mitigate natural gas transportation and supply issues in extreme weather events, such as hurricanes. FRCC s Fuel Reliability Working Group (FRWG) provides oversight of the Regional Entity fuel reliability forum that studies fuel availability and coordinates responses to fuel issues and emergencies. MISO Policy and changing generation trends continue to drive new potential risks to resource adequacy and will require continued transparency and vigilance to ensure long-term needs. MISO projects that reserve margins will continue to tighten over the next five years, which approaches the reserve margin requirement. Operating at the reserve margin creates a new operating reality for MISO members where the use of all resources available on the system and emergency operating procedures are more likely. This reality will lead to a projected dependency in use of load modifying resources, such as behind-the-meter generation and demand response (DR). A number of large resources continue to feel economic pressure, which could lead to further plant retirements and drive the reserve margin lower. MRO-Manitoba Hydro There are potential new electricity export opportunities between Manitoba and Saskatchewan, which would likely require new transmission in western Manitoba. There is uncertainty as to the availability of local voltage support due to the potential shutdown of Brandon Unit 5 in 2020 in western Manitoba. It is expected that post-disturbance voltage will become an emerging concern in the period beyond 2025 depending on the timing of various projects. Plans to address this are under study. MRO-SaskPower The requirement to reduce emissions from thermal generating facilities will call for ongoing planning to ensure that proposed thermal generation retirements are successfully implemented. Saskatchewan is also working with the provincial and federal governments on emission regulations and agreements to confirm the schedule for retirements. Saskatchewan will have an increase in wind generation in the near- and long-term planning horizons. The inclusion of more intermittent resources may have operational impacts such as changes to net demand ramping variability that need to be studied to determine the power system effects on both Saskatchewan and neighboring jurisdictions. 71

81 Chapter 6: Regional Overview NPCC-Maritimes The Maritimes Area has begun tracking the ramp rate variability trend but does not yet have enough years of data for the area as a whole to identify any trends. Given the essentially flat load growth and small degree of anticipated variable energy resource (VER) installations, little change in either ramp rates or the area s resource mix is expected to occur for the duration of the LTRA assessment period. The maximum net demand ramping variability 1 hour up, 1 hour down, 3 hours up, and 3 hours down values for two historical years of 2014 and 2015 and a future year of 2020 were calculated along with the percentage contributions of VERs versus the loads. The majority of the maximums occurred during the late fall shoulder and winter peak seasons NPCC-New England Solar PV resources constitute the largest segment of distributed generation resources throughout New England. The region has experienced significant growth in the development of PV resources over the past few years and continued growth of PV is anticipated. In order to determine what impacts future PV could have on the regional power grid, the ISO created a forecast of future PV. ISO-NE s solar forecast separates the PV into two categories: 1) Markets and 2) Behind-the-Meter. PV in the Markets category consists of resources participating in the forward capacity market and PV as energy-only generating assets that participate in the energy market. Behind-the-Meter PV comprises approximately two-thirds of the total PV capacity and is treated as a load reducer. ISO-NE has limited information on the characteristics of behind-the-meter PV resources. ISO-NE does not collect behind-the-meter PV metered data, but can estimate its operational characteristics by using available historical PV production data along with total installed nameplate capacity. The total peak load reduction value of all PV in New England amounted to 588 MW in 2016 and it is forecasted to grow to 964 MW by 2021 and to 1,127 MW by These summer peak load reduction values are calculated as a percentage of ac nameplate. The percentages, which include the effect of diminishing PV production as increasing PV penetrations shift the timing of peaks later in the day, decrease from 40 percent of nameplate in 2015 to about 34 percent in The ISO surveys Distribution Owners several times a year to determine the historical installations of PV and other types of distributed generation. The ISO annually projects PV installations by state and distributes them by dispatch zone. The forecast is based on state policies and reflects inputs from stakeholders. With the exception of PV, distributed generation is growing slowly and is accounted for within the ISO s demand forecast. The ISO is currently conducting scenario analyses that reflect large scale development of PV. These long-term planning studies, scheduled for completion in 2017, will assess the potential impacts on operating reserves, ramping, and regulations. NPCC-New York On January 25, 2016, the New York State Department of Public Service Staff (DPS) issued a whitepaper outlining its recommendations to the NYSPSC for implementing the state s Clean Energy Standard (CES). 71 The CES is intended to increase the amount of renewable energy generation in New York State to 50 percent of total generation by 2030 while retaining upstate nuclear power plants in support of the state s carbon dioxide emissions reduction goals. The current solar integration study concluded that the BPS can reliably manage (over the five-minute time horizon) the increase in net load variability associated with the solar PV and wind penetration levels up to 4,500 MW wind and 9,000 MW solar PV. The solar study also concluded that the large-scale implementation of behind-the-meter solar PV will impact NYISO s load profile and associated system operations. Also, the lack of frequency and voltage ride-through requirements for solar PV facilities could worsen system contingencies when solar PV deactivates in response to frequency and voltage excursions 71 New York Department of Public Service: Staff White Paper on Clean Energy Standard; January

82 Chapter 6: Regional Overview New York only determines an annual installed reserve margin (IRM) to meet a one-day-in-ten loss of load expectation (LOLE). Estimating the impact of the above-referenced issues on the IRM is difficult due to so many different input variables that would increase or decrease the margins. However, the addition of intermittent resources, such as solar and wind in the amounts proposed by certain initiatives, would have the tendency to increase the IRM requirements over time. NPCC-Ontario With the growth in distributed generation capacity, demand forecasting has become increasingly more complex. Traditionally, demand was mainly a function of weather conditions, economic cycles, and population growth. With multiple new factors influencing demand, such as increased distribution-connected VER and increased consumer price-responsiveness, determining the causality of demand changes has become increasingly nuanced. The introduction of VERs (e.g., solar and wind), the removal of flexible generation (coal), and lower demand and limitations in operational flexibility of gas and hydro resources have added new challenges to maintaining a reliable system. The results of a recent operability assessment indicated that there is a system need for enhanced flexibility to balance supply and demand, more regulation, and additional grid voltage control. It is important that the supply mix remains robust in meeting industry planning standards, flexible to meet the ever-changing demands of system operations, and balanced to manage inherent risks (e.g., fuel security and critical infrastructure needs). To that end, the IESO has launched an initiative to augment resource flexibility and issued a Request for Information for additional regulation service in June The IESO has an energy storage pilot program underway to test the capability of storage technologies to provide grid services as well. Activities are also underway with transmitters to plan and install additional dynamic and static voltage control devices to help with voltage control. Increasing amounts of VERs and relatively flat demand levels have contributed to a rise in surplus baseload generation (SBG) in Ontario. Over the next few years, more VERs are expected, but the effects on SBG will be tempered by the impact of the planned nuclear refurbishments and retirements. The IESO has mechanisms in place to manage SBG, including economic exports, wind and solar dispatch, and nuclear maneuvers or shutdowns. NPCC-Québec While technical developments in recent years have contributed to build a more reliable system, sustainable system reliability may be challenged by emerging issues, such as potential operational issues due to the changing resource mix. In the Québec area, wind generation capacity has increased by 2,500 MW over the last five years, but the area s total installed capacity is still mainly composed of large reservoir hydro complexes (more than 90 percent) that can react quickly to adjust their generation output and meet the sharp changes in electricity net demand. The forecasted change to resource mix is not expected to have any influence on the ramp rate trends or any other reliability issue. PJM PJM has experienced some thermal overload problems during light load conditions with relatively high wind generator output. PJM s light load reliability analysis ensures that the transmission system is capable of delivering the system generating capacity at light load. The system generating capability modeling assumption for this analysis is that the generation modeled reflects generation by fuel class that historically operates during the light load demand level, such as high wind output. Interchange levels for the various PJM zones will reflect a statistical average of typical previous years interchange values for off-peak hours. Load level, interchange, and generation dispatch for non-pjm areas impacting PJM facilities are based on statistical averages for previous off-peak periods. The flowgates ultimately used in the light load reliability analysis are determined by the following: running all contingencies maintained by PJM planning and monitoring all PJM market-monitored facilities and all BPS facilities. The contingencies used for light load 73

83 Chapter 6: Regional Overview reliability analysis will include NERC TPL P1, P2, P4, P5, and P7. For NERC TPL P0, normal system conditions will also be studied. SERC With respect to MISO, a settlement agreement was reached between MISO, SPP, and the Joint Parties (TVA, SOCO, LG&E/KU, AECI and PowerSouth). The settlement agreement is now in effect (and superseded the ORCA on February 1, 2016) to reliably manage the magnitude of power transfers between MISO South and Midwest. The settlement agreement limits transfers between MISO-South and MISO-Midwest to 2,500 MW and between MISO- Midwest to MISO-South to 3,000 MW in order to limit reliability impacts on neighboring systems. The increase in flow from 1,000 to 2,500/3,000 MW represents a new operating condition that has been studied and experienced under certain historical operating conditions. However, this is a significant change that will be closely monitored in operations for adverse reliability impacts. Although a settlement agreement is in place, SERC is committed to ensuring reliability of the region and the interconnection. The region implemented a Joint Loop flow study initiative with market and nonmarket entities to recreate and study loop flows within the area. The purpose of these studies are to ensure there are not potential IROL conditions that can lead to cascading, separation, or blackout conditions SPP SPP, along with other joint parties in the Region, and MISO are currently managing reliability concerns from MISO s recent operational changes under the provisions of the Operations Reliability Coordination Agreement (ORCA). 72 On March 1, 2015, SPP and MISO began using Market-to-Market mechanisms to more efficiently and economically control congestion on SPP and MISO flowgates, in which both markets have a significant impact. During congestion on an SPP market-to-market flowgate, SPP will initiate the market-to-market process, and SPP and MISO will coordinate through an iterative process to identify and dispatch the most cost-effective generation between the two markets to relieve the congestion. Texas RE-ERCOT The Texas panhandle region is currently experiencing significantly more wind generation developer interest than what was initially planned for the area. ERCOT and Texas-RE are conducting a study while participating in a recent NERC pilot project related to ramping issues associated with high levels of VERs. In addition, ERCOT is performing steady state, dynamic, and short-circuit assessments to identify weak system areas and in particular to assess the reliability impacts in areas with high renewable penetrations. The assessment area is also projecting potential high increases in small distributed generation additions, such as rooftop solar. ERCOT is accelerating its efforts to more accurately map distributed energy resources (DERs) to the transmission grid. WECC Load-serving entities historically experience two rapid increases in customer demand: early morning and late afternoon. These rapid changes were typically balanced by increased hydroelectric and thermal generation. However, with greater generation contribution of intermittent resources, hydro and thermal units are required to follow larger daily demand fluctuations. The CA/MX subregion is seeing a large increase in distributed resources. There is currently about 4,300 MW of rooftop solar installed in the Western Interconnection with about 4,000 MW of that total installed in the CA/MX subregion. By 2026, that total is expected to increase to over 12,000 MW in the interconnection with over 11,000 MW installed in the CA/MX subregion. Due to the importance of hydro generation from the northwest, WECC monitors hydro conditions in that region. Under the Pacific Northwest Coordination Agreement, entities within the Northwest PowerPool have the obligation to coordinate the operations and long-term planning for the northwest Hydro system. WECC relies on their obligation and expertise to monitor and manage NW hydro issues. 72 Operations Reliability Coordination Agreement; June

84 Chapter 6: Regional Overview California is continually working to coordinate the integration of VERs. The California ISO s energy imbalance market (EIM) provides a way to share energy reserves and renewable energy through a real-time energy market. The EIM operates in parts of California, Nevada, Utah, Idaho, Washington, Oregon, and Wyoming, and will expand into parts of Arizona in The ability to dispatch resources throughout the Western Interconnection has increased the flexibility needed to incorporate VERs into the grid. Probability-Based Resource Adequacy Assessment NERC recognizes that a changing resource mix with significant increases in energy-limited resources, changes in off-peak demand, and other factors can have an effect on resource adequacy. As a result, NERC is incorporating more probabilistic approaches into this assessment as well as other ongoing analyses that will provide further insights into how to best establish adequate reserve margins amidst a BPS undergoing unprecedented changes. Historically, NERC has gauged resource adequacy through planning reserve margins, which are deterministic assessment metrics. Planning reserve margins are a measure of available capacity over and above the capacity 73, 74 needed to meet normal (50/50) forecast peak demand. Background In 2010, the Generation and Transmission Reliability Planning Models Task Force (GTRPMTF) concluded that existing reliability models could be used to develop one common composite generation and transmission assessment. The task force also noted the importance of having complete coverage of the North American BPS as well as the elimination of overlaps. As this premise is already adopted and executed annually in the LTRA, the approach for the probabilistic assessment follows suit. The assessment areas (i.e., Regions, Planning Coordinators (PCs), independent system operators (ISOs), and regional transmission organizations (RTOs)) used for this assessment are identical to those used for the LTRA. NERC produced a series of probabilistic assessment reports conducted by the Regions and assessment areas, covering all of the NERC Assessment Areas. In this effort to improve NERC s continuing probabilistic and deterministic assessments, the Probabilistic Assessment Improvement Task Force 1 (PAITF) was formed in May, 2015, from members of the Planning Committee (PC), the Reliability Assessment Subcommittee (RAS), and selected observers from industry. Its purpose is to support the identification of improvement opportunities for NERC s Long-Term Reliability Assessment and complementary probabilistic analysis. PAITTF has developed two reports. The first is the NERC Probabilistic Assessment Improvement Plan report, published in December This report provided possible recommendations by PAITF based on recent LTRA key findings for NERC core and proposed coordinated special probabilistic assessment reports. The second report was the NERC Technical Guideline document published in August, This report provided detailed probabilistic modeling guidelines and technical recommendations that serve as a platform for detailing probabilistic analytical enhancements that apply to resource adequacy. 73 NERC Reliability Assessments 74 NERC Probabilistic Assessment Improvement Task Force 75

85 Chapter 6: Regional Overview In the development of the NERC 2016 Probabilistic Assessment, NERC RAS and RAS-ProbA team implemented the following main PAITF technical guideline recommendations: Regions and assessment areas calculate monthly resource adequacy metrics. As resource and demand characteristics change over time, annual loss of load may start accruing during historically off-peak months. Therefore, the monthly aggregation of these metrics [loss of load hours (LOLH) and expected unserved energy (EUE)] will better inform industry of potential resource adequacy risks throughout the year. Assessment areas performed sensitivity modeling within the core probabilistic assessment framework. The NERC RAS identified the variable data elements relevant to each sensitivity model. NERC, with input from the RAS, ERO-RAPA, and the Planning Committee (PC), identified the Sensitivity Case to be an increase in load growth for the 2016 core probabilistic assessment. The purpose of this Sensitivity Case is to demonstrate the robustness of the loss of load measures: Increase on-peak demand by two percent in the second study year, 2018, and by four percent in the fourth study year, Increase MWh net energy by two percent in second study year, 2018, and by two percent in the fourth study year, Summary statistic results of the forecast planning reserve margin, the forecast operable reserve margin, 75 annual and monthly LOLH and EUE measures for the Base Case and the Sensitivity Case, and high-level key findings are presented in the Assessment Area Granular Review section of the report. Assessment Area Granular Review Provided in more detail here are the more granular reviews of each assessment area s footprint, methods, assumptions used for this assessment, and additional information and data. This section includes dashboards that review both the deterministic and probabilistic data and results. Individually, these present a straightforward overview of the different assessment areas that make up the BPS and the variations between them. 75 Forecast Operable Reserve Margin is defined as the ratio of anticipated resources derated by forced outage rates less on peak demand. 76

86 Chapter 6: Regional Overview FRCC The Florida Reliability Coordinating Council s (FRCC) membership includes 30 Regional Entity Division members and 23 Member Services Division members composed of investor-owned utilities (IOUs), cooperative systems, municipal utilities, power marketers, and independent power producers. FRCC is divided into 10 Balancing Authorities with 47 registered entities (both members and nonmembers) performing the functions identified in the NERC Reliability Functional Model and defined in the NERC Reliability Standards. The Region contains a population of over 16 million people and has a geographic coverage of about 50,000 square miles over Florida. Summary of Methods and Assumptions Reference Margin Level The FRCC Region utilizes the NERC 15 percent reference margin. Load Forecast Method Noncoincident, based on individual forecasts Peak Season Summer Planning Considerations for Wind Resources No wind capacity Planning Considerations for Solar Resources Small amount of solar capacity; based on historical average at peak Footprint Changes Region is the assessment area footprint; no recent changes Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 48,125 48,648 49,266 49,873 50,461 50,973 51,514 52,125 52,803 52,803 Demand Response 3,014 3,070 3,123 3,167 3,205 3,255 3,271 3,271 3,304 3,304 Net Internal Demand 45,111 45,578 46,143 46,706 47,256 47,718 48,243 48,854 49,499 49,499 Resources (MW) Anticipated 55,015 55,019 57,442 58,088 58,379 58,551 58,916 60,533 60,994 60,976 Prospective 55,436 55,440 58,024 58,656 59,445 59,657 60,062 62,405 62,981 62,978 Reserve Margins (%) Anticipated 21.95% 20.71% 24.49% 24.37% 23.54% 22.70% 22.12% 23.91% 23.22% 23.19% Prospective 22.89% 21.64% 25.75% 25.59% 25.79% 25.02% 24.50% 27.74% 27.24% 27.23% Reference Margin Level 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Shortfall (MW) Anticipated Prospective

87 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: FRCC used the Tie Line and Generation Reliability (TIGER) program, which is based on the analytical method of recursive convolution for the computation of LOLH and EUE metrics. Modeling: The current modeling approach incorporates regional hourly load, generation data, forced outage rates, maintenance schedules, and monthly DR. Additionally, a load variation model was utilized that provided 500 variations of annual hourly load as an input into TIGER. FRCC was modeled as an isolated area with no interconnections with other areas and allowing only firm imports. Results Trending: 2018 was studied in both the 2014 and the 2016 ProbA to evaluate any changes or trends. The 2014 ProbA Base Case analysis resulted in an EUE of MWh and an LOLH of hours per year. The results from the 2016 ProbA Base Case analysis showed a negligible decrease. Probabilistic vs. Deterministic Reserve Margin Results: There are no differences between the reserve margin reported in the LTRA and ProbA Base Case. Base Case Study Reserve margins for the study years are well above the NERC reference margin of 15 percent, resulting in low LOLH and EUE values. The EUE was MWh in 2018 and MWh in Projected loss of load only occurred during the summer season. Sensitivity Case Study With the increase of load in the Sensitivity Case, reserve margins remain above the NERC reference margin of 15 percent, and the EUE increased slightly from the Base Case to MWh in 2018 and MWh in Similar to the Base Case, a nonzero loss of load values are projected only during the summer season with highest values in August. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

88 Chapter 6: Regional Overview Planning Reserve Margins, Demand FRCC has a reliability criterion of a 15 percent minimum regional total reserve margin based on firm load. FRCC reserve margin calculations include merchant plant capacity that is under firm contract to load-serving entities. FRCC assesses the upcoming ten-year projected summer and winter peak hour loads, generating resources, and demand-side management (DSM) resources on an annual basis to ensure that the regional reserve margin requirement is projected to be satisfied. The three Florida Investor Owned Utilities, Florida Power & Light Company (FPL), Duke Energy Florida (DEF), and Tampa Electric Company (TEC) are utilizing, along with other reliability criteria, a 20 percent minimum total reserve margin planning criterion consistent with a voluntary stipulation agreed to by the Florida Public Service Commission (FPSC). 76 Other utilities employ a 15 percent to 18 percent minimum total reserve margin planning criterion. 77 Based on the expected load and generation capacity, all projected reserve margins are above the NERC Reference Margin Level of 15 percent for the FRCC assessment area with FRCC reserve margins remaining above 20 percent for all seasons during the assessment period. 76 Docket No EU Generic investigation into the aggregate electric utility reserve margins planned for Peninsular Florida, Order No. PSC S-EU, issued December 22, 1999 ( 77 Florida Administrative Code Rule Adequacy of Resources ( 79

89 Chapter 6: Regional Overview FRCC continues to project growth in peak load, but the projected growth is less than in the previous forecast. The net energy for load (NEL) and summer peak demands are forecasted to be lower than in the previous forecasts. The current average annual growth rate for NEL is 0.8 percent per year compared to 1.1 percent per year in the previous forecast. Firm summer peak demand is expected to grow by 1.1 percent per year compared to 1.5 percent peak demand growth rate in the previous forecast. This is primarily due to more utilities starting to capture appliance efficiency in their load forecast models or using updated appliance efficiency assumptions. For firm winter peak demand, the average growth rate is now expected to be 1.0 percent per year compared to 0.9 percent per year in the previous forecast. Demand-Side Management The FRCC Region is projecting some decrease in the growth rate of utility program energy-efficiency due to two factors: (1) significant decreases in DSM cost-effectiveness caused by lower fuel costs, etc. and (2) increased impacts from federal and state energy-efficiency codes and standards (e.g., 2005 National Energy Policy Act, 2007 Energy Independence and Security Act). The impacts from these energy-efficiency codes and standards is lowering the potential for utility energy efficiency programs to lower demand and energy usage for appliances and equipment addressed by the codes and standards. However, these codes and standards are resulting in significant reductions in demand and energy that are accounted for in load forecasts. DR from interruptible and load management programs within FRCC is treated as a load-modifier and is projected to be relatively constant at approximately 6.4 percent of the summer and winter total peak demands for all years of the planning horizon. The Florida Public Service Commission (FPSC) evaluates and revises its DSM goals every five years. New DSM Goals were set in Because of diminished cost-effectiveness of DSM programs, and the fact that energy-efficiency codes and standards have lowered the potential for DSM programs, the FPSC set lower DSM Goals for Florida utilities than had been previously set in DR from interruptible and load management programs within FRCC is treated as a load-modifier, and is projected to be relatively constant at approximately 6.4 percent of the summer and winter total peak demands for all years of the planning horizon. Generation FRCC is projecting approximately 12,000 MW of summer and 12,262 MW of winter Tier 1 capacity to be added during the assessment period. The Tier 1 capacity will consist of mainly natural gas capacity with approximately 300 MW of firm solar (PV) and 180 MW of biomass. There are also 548 MW of planned uprates during the assessment period. The proposed generation additions are studied by the interconnecting Transmission Owner as well as by the FRCC Transmission Working Group (TWG) through FRCC s Transmission Service and Generator Interconnection Service Request Assessment Area Deliverability Evaluation Process. 79 Entities within FRCC have capacity transfers that have firm contracts available to be imported into the assessment area from SERC. There is approximately 830 MW of FRCC member-owned generation that is dynamically dispatched out of the SERC assessment area. These imports have firm transmission service to ensure deliverability into the FRCC assessment area. All firm on-peak capacity imports into the FRCC Region have firm transmission service agreements in place to ensure deliverability into the FRCC Region with these capacity resources included in the calculation of the Region s Anticipated Reserve Margin. In addition, the interface owners between FRCC and SERC assessment areas meet quarterly to coordinate and perform joint studies to ensure the reliability and adequacy of the interface. The FRCC assessment area is projecting approximately 3,900 MW of summer generation to be retired through the assessment period. These retirements will include approximately 2,400 MW of natural gas generation, 1000 MW of coal, and 500 MW of oil. Also, a 400 MW natural gas unit will be converted to a synchronous condenser to 78 Order No. PSC FOF-EU ( 79 FRCC Reliability Evaluation Process For Generator and Transmission Service Requests; FRCC-MS-PL-054;

90 Chapter 6: Regional Overview provide voltage support. Based on the annual FRCC long-range study, FRCC is not anticipating any reliability impacts resulting from these unit retirements, which are studied as part of the FRCC long-range study process performed annually by the TWG to mitigate potential reliability impacts to the Grid and the FRCC reserve margin criteria. FRCC is not anticipating any larger generator unavailability during system peak. All known scheduled generation outages in the long-term horizon are incorporated into the annual FRCC long-range study process to mitigate any potential reliability impacts to the BPS. FRCC s Fuel Reliability Working Group (FRWG) provides oversight of the Regional fuel reliability forum that studies the fuel availability and coordinates responses to fuel issues and emergencies. FRCC is not expecting any longterm reliability impacts resulting from an increase mixture of natural-gas-fired generation. Transmission and System Enhancements The FRCC Region has not identified any major projects that are needed to maintain or enhance reliability during the planning horizon. Planned projects are primarily related to expansion in order to serve forecasted growing demand and maintain the reliability of the BPS in the longer- term planning horizon. The FRCC Region is not anticipating any additional reliability impacts resulting from potential environmental regulations. The State of Florida has not developed a renewable portfolio standard establishing target renewable thresholds. The 2013 MATS study performed by the FRCC s Transmission Working Group identified reliability impacts resulting from the retirement of two coal units at the same site. These units were granted an extension and will be able to run through These units will be replaced in the year 2018 by two gas-fired combined cycle units to maintain the reliability of the BPS within the FRCC Region. However, FRCC will continue to monitor the progress of the Clean Power Plan (CPP) to determine the potential impact to reliability once the current legal challenge has been resolved, which may result in changes in the timing and/or substance of the current CPP final rules. Long-Term Reliability Issues FRCC has not identified any long-term reliability issues. The FRCC Region performed an extreme weather (105 percent of peak) sensitivity scenario into its 2015 annual long-range planning study process to identify any potential reliability impacts to the BPS. The FRCC region is not expecting any reliability impacts during the shoulder periods. For the FRCC region, the shoulder periods are the spring and fall seasons. These seasons are studied in the operational horizon by the Operational Planning Working Group (OPWG) and by the TWG in the long-term horizon (off-peak cases) as part of the annual long range assessment. Additionally, FRCC has not identified any other emerging reliability issues. However, FRCC continues to monitor the possible impacts on the long-term reliability of the BPS from pending environmental legislations (CSAPR, NESHAP, RICE, MATS, and CPP). 81

91 Chapter 6: Regional Overview MISO The Midcontinent Independent System Operator, Inc. (MISO) is a not-for-profit member-based organization. MISO administers wholesale electricity markets that provide customers with valued service, reliable, cost-effective systems and operations, dependable and transparent prices, open access to markets, and planning for long-term efficiency. MISO manages energy, reliability, and operating reserve markets that consist of 36 local Balancing Authorities and 394 market participants, serving approximately 42 million customers. Although parts of MISO fall in three NERC Regions, MRO is responsible for coordinating data and information submitted for NERC s reliability assessments. Summary of Methods and Assumptions Reference Margin Level 15.2 percent This increase is mainly driven by a process change within the LOLE study. Load Forecast Method Coincident Peak Season Summer Planning Considerations for Wind Resources Effective load-carrying capability (ELCC); varies by wind node Planning Considerations for Solar Resources No utility-scale solar resources in MISO Footprint Changes Minnesota is reporting under MISO this year Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 127, , , , , , , , , ,462 Demand Response 5,827 5,827 5,827 5,827 5,827 5,827 5,827 5,827 5,827 5,827 Net Internal Demand 121, , , , , , , , , ,635 Resources (MW) Anticipated 143, , , , , , , , , ,297 Prospective 150, , , , , , , , , ,047 Reserve Margins (%) Anticipated 18.09% 17.50% 17.63% 16.60% 15.97% 13.89% 12.17% 11.54% 9.83% 9.07% Prospective 23.78% 23.71% 24.71% 26.85% 26.17% 23.89% 22.21% 21.53% 19.81% 18.98% Reference Margin Level 15.20% 15.20% 15.20% 15.20% 15.20% 15.20% 15.20% 15.20% 15.20% 15.20% Shortfall (MW) Anticipated ,640 3,836 4,651 6,862 7,890 Prospective

92 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: MISO is a summer-peaking system that spans 15 states and consists of 36 local Balancing Authorities that are grouped into 10 local resource zones. For the ProbA, MISO utilized a multiarea modeling technique for the 10 local resource zones internal to MISO. Firm external imports and nonfirm imports are also modeled. This multi-area modeling technique for resource zones and accompanying methodology has been thoroughly vetted through MISO s stakeholder process. Modeling: Each local resource zone was modeled with an import and export limit based on power flow transfer analysis. In addition to the zone-specific import and export limits, a regional directional limit was modeled that limits the Midwest (local resource zones 1 7) to south (local resource zones 8 10) flow to 3,000 MWs and the south to Midwest to 2,500 MWs. The modeling of this limit is the main driver for the difference between the probabilistic and deterministic reserve margins. MISO utilizes unit specific outage, planning, and maintenance outage rates within the analysis based on five years of Generation Availability Data System (GADS) data. Modeling unit specific outage rates increases precision in the probabilistic analysis when compared to the utilization of class average outage rates. Results Trending: Previous results in the 2014 Probabilistic Assessment resulted in MWh EUE and 0.09 Hours per year LOLH. The results from this year s analysis resulted in a slight decrease for 2018 when compared to the analysis completed in the 2014 Probabilistic Assessment. Probabilistic vs. Deterministic Reserve Margin Results: The LTRA deterministic reserve margins decrease capacity that is constrained within MISO south due to the 2,500 MW limit which reflects a decrease in reserve margin. The constraint was explicitly modeled for the probabilistic analysis and determined if sufficient capacity was available to transfer from south to north and vice versa. The modeling of this limitation produces an increase for the ProbA forecast planning reserve margin. 83

93 Chapter 6: Regional Overview Base Case Study The bulk of the EUE and LOLH are accumulated in the summer peaking months with some off-peak risk. Increases in loss of load statistics are expected with decreasing reserve margins. Sensitivity Case Study The Sensitivity Case is a good proxy for increased retirement risk and/or increased load forecasts. The percent increase is equal to a 2,565 MW increase and the percent increase is equal to a 5,203 MW increase. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

94 Chapter 6: Regional Overview Overview MISO projects a regional surplus for the summer of 2017 with potential regional shortfall starting in These results show a potential regional short fall two years earlier than the 2015 MISO LTRA results. These results are driven by a number of factors: A decrease in resources committed to serving MISO s load mainly by independent power producers (IPP). A decrease in load forecasts where the biggest drop was in Zone 6 (Indiana). The increase in committed resources (Tier 1) in Zone 7 (Michigan). MISO projects that each zone within the MISO footprint will have sufficient resources within their boundaries to meet their local clearing requirements or the amount of their local resource requirement (which must be contained within their boundaries). Several zones are short against their total zonal reserve requirement when only resources within their boundaries (or are contracted to serve their loads) are considered. However, those zones have sufficient import capability, and the MISO region has sufficient surplus capacity in others zones to support this transfer. Surplus generating capacity for zonal transfers within MISO could become scarce in later years if no action is taken in the interim by MISO load-serving entities. All zones within MISO are sufficient from a resource adequacy point of view in the near term when available capacity and transfer limitations are considered. Regional shortages in later years may be rectified by the utilities; MISO is engaged with stakeholders in a number of resource adequacy reforms to help rectify these later year s shortages. Policy and changing generation trends continue to drive new potential risks to resource adequacy, requiring continued transparency and vigilance to ensure long-term needs. MISO projects that reserve margins will continue to tighten over the next five years and approach the reserve margin requirement. Operating at the reserve margin creates a new operating reality for MISO members where the use of all resources available on the system and emergency operating procedures are more likely. This reality will lead to a projected dependency in use of Load Modifying Resources, such as behind-the-meter generation and DR. The SPP settlement agreement has put in place a Regional Directional Transfer Limit replacing the ORCA operating limit. Specifically the Midwest (LRZs 1-7) to south (LRZs 8-10) flow is limited to 3,000 MWs and south to Midwest is limited to 2,500 MWs. 80 This year marks the third iteration of the Organization of MISO States (OMS) MISO survey, which helps provide forward visibility into the resource adequacy position of the MISO region. The survey also helped identify resources that had a low certainty of being available for each planning year. The LTRA results represent a point in time forecast, and MISO expects these figures will change significantly as future capacity plans are solidified by load-serving entities and States. For example, there are enough resources in Tier 2 and 3 to mitigate any long-term resource shortfalls. MISO forecasts the coincident Total Internal Demand to peak at 127,607 MW during the 2017 summer season. This is a decrease of roughly 2,700 MWs from last year s projection for This decrease is mainly driven by load reductions in Zones 5 (Missouri) and aluminum smelter closures in Zone 6 (Indiana). MISO projects the 80 MISO Presentation: SPP Settlement Update; October

95 Chapter 6: Regional Overview summer coincident peak demand to grow at an average annual rate of 0.6 percent, which is less than the growth rate from the 2015 assessment. As a result of the OMS-MISO survey, resources with a low certainty of being available for the given year are more visible. This number is small in Years 1 3 and then ramps up in the future. The reductions of these low certainty resources are more than offset with Tier 2 and 3 resources and should not cause any resource adequacy issues. However, MISO continues to see a number of large resources, generally IPPs, that are at-risk for retirement due to economics. Local reliability issues could result with some of the unannounced retirements. The annual MISO Transmission Expansion Plan (MTEP) 81 proposes transmission projects to maintain a reliable electric grid and deliver the lowest-cost energy to customers in the MISO region. As part of MTEP15, MISO staff recommends $2.75 billion of new transmission expansion through 2024, as described in Appendix A of the MTEP report, 82 to the MISO Board of Directors for review, approval, and subsequent construction. The 345 new projects in MTEP15 Appendix A represent $2.75 billion 83 in transmission infrastructure investment and fall into the following four categories: 90 Baseline Reliability Projects (BRP) totaling $1.2 billion: BRPs are required to meet NERC reliability standards. 12 Generator Interconnection Projects (GIP) totaling $73.6 million: GIPs are required to reliably connect new generation to the transmission grid. 1 Market Efficiency Project (MEP) totaling $67.4 million: MEPs meet requirements for reduction in market congestion. 242 Other Projects totaling $1.38 billion: Other projects include a wide range of projects, such as those that support lower-voltage transmission systems or provide local economic benefit, but do not meet the threshold to qualify as Market Efficiency Projects. MISO is working with stakeholders to create resource adequacy reforms to move to a seasonal construct. The seasonal construct would create a summer and a winter planning reserve margin requirement and seasonal resource parameters (on peak capacity, EFORd, etc.). The seasonal construct will better reflect the seasonality of the wind, solar, etc. and increase the visibility of reliability in the winter season. 81 MISO Transmission Expansion Planning (MTEP) 82 MISO Transmission Expansion Plan The MTEP15 report and project totals reflect all project approvals during the MTEP15 cycle, including those approved on an out-of-cycle basis prior to December

96 Chapter 6: Regional Overview MRO-Manitoba Hydro Manitoba Hydro is a Provincial Crown Corporation that provides electricity to 561,869 customers throughout Manitoba and natural gas service to 274,817 customers in various communities throughout southern Manitoba. The province of Manitoba is 250,946 square miles. Manitoba Hydro is winter peaking. No change in the footprint area is expected during the assessment period. Manitoba Hydro is its own Planning coordinator and Balancing Authority. Manitoba Hydro is a coordinating member of the MISO. MISO is the Reliability Coordinator for Manitoba Hydro. Summary of Methods and Assumptions Reference Margin Level The capacity criterion, as determined by Manitoba Hydro, requires a minimum 12 percent planning reserve margin, applied as the Reference Margin Level in this assessment. Load Forecast Method Coincident Peak Season Winter Planning Considerations for Wind Resources Effective Load-Carrying Capability (ELCC) of 15.6 percent for the summer and 20 percent for the winter. Planning Considerations for Solar Resources No utility-scale solar resources Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 4,826 4,713 4,646 4,703 4,685 4,710 4,743 4,781 4,793 4,821 Demand Response Net Internal Demand 4,826 4,713 4,646 4,703 4,685 4,710 4,743 4,781 4,793 4,821 Resources (MW) Anticipated 5,419 5,557 5,647 6,304 6,412 6,437 6,437 6,437 6,412 6,412 Prospective 5,526 5,649 5,646 5,961 5,844 5,869 5,869 5,869 5,969 5,969 Reserve Margins (%) Anticipated 12.29% 17.90% 21.53% 34.04% 36.86% 36.66% 35.72% 34.65% 33.79% 33.00% Prospective 14.51% 19.86% 21.53% 26.75% 24.74% 24.61% 23.76% 22.77% 24.56% 23.82% Reference Margin Level 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% Shortfall (MW) Anticipated Prospective

97 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: Manitoba Hydro system is a winter-peaking system, and the vast majority of its generating facilities are use-limited or energy-limited hydro units. The 2016 Manitoba Hydro probabilistic assessment was conducted using the Multi-Area Reliability Simulation (MARS) program. The data used in the MARS simulation model are consistent with the data reported in the 2016 LTRA submittals from Manitoba Hydro to NERC. Modeling Characteristics: Manitoba Hydro and its neighboring systems are modeled as two areas consisting of Manitoba and the northwest part of MISO. Each of the two interconnected areas are modeled connected directly and the transmission between Manitoba and MISO is modeled with interface transfer limits. Three different types of resources are modeled for Manitoba Hydro system: hydro resources, thermal resources (including both coal and gas units), and intermittent wind resources. The 8,760 point hourly load records of a typical year were used to model the annual load curve shape. Load forecast uncertainty is modeled in both the Base and Sensitivity Cases. DR programs are modeled as a simple load modifier by reducing the peak load. Contractual obligations are modeled as load modifiers considering the contractual obligations of the power sales and purchase agreements. Results trending: The LOLH and EUE values obtained in the 2014 Probabilistic Assessment are zero. The nonzero LOLH and EUE values are obtained for both the Base and Sensitivity cases in 2016 Probabilistic Assessment. The slight increase in the reliability indices is mainly due to the changes in modeling assumptions. The following specific changes are made in 2016 assessment as compared to 2014 assessment: 1) Multiple flow conditions, including an extreme drought scenario, are modeled and the indices calculated are weighted averages of the indices obtained for different water conditions. 2) Increased standard deviation of the seven-step load forecast uncertainty from four percent to five percent. 88

98 Base Case Study For 2018 Base Case, small values of EUE and LOLE are observed due to a relatively smaller reserve margin. For 2020 Base Case, the reserve margin is increased significantly due to the expected addition of a new generating station and therefore the LOLE and EUE are virtually zero. All loss of load events are in winter season and the highest contribution to loss of load is from the winter month of November. Sensitivity Case Study As expected, the reliability indices are increased in the Sensitivity Cases for both the 2018 and 2020 planning years, and all loss of load events Chapter 6: Regional Overview Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) are in winter season. Although the planning reserve margin drops below the reference value of 12 percent for a 2 percent increase in peak load, the EUE and LOLE are still small for 2018 planning year. The minor changes in the LOLE and EUE indices for 2020 planning year is mainly due to the decrease in reserve margin for a 4 percent increase in peak load. The highest contribution to the loss of load event is still from the winter month of November for 2018 while it is from the winter month of March for 2020 planning year. 89

99 Chapter 6: Regional Overview Demand, Resources and Reserve Margins Manitoba Hydro is projecting reserve margins above the Reference Margin Level during the assessment period. Since the previous assessment, the demand forecast is projected to be 1.4 percent lower by 2025/26, and this is primarily attributable to an increase in the forecast of DSM activity. The forecast of DSM in 2025/26 is increasing from 516 MW forecast in the previous assessment to 685 MW in the current assessment. No changes were made to the load forecasting methodology from the last assessment period. Energy efficiency and conservation savings are forecast higher than prior year s assessment due to enhancements to existing programs (e.g., increasing program incentives, adding alternative program delivery methods, or adding new measures to the program). There have been no capacity additions in Manitoba since the 2015 LTRA. The Keeyask Hydro Generating Station is now under construction and is considered a Tier 1 capacity addition. Manitoba Hydro is anticipating the first units of the 630 MW of net capacity addition from the Keeyask Hydro Generating Station would begin to come into service in late Brandon Unit 5, Manitoba Hydro s sole remaining coal-fired generating unit, is assumed to remain available until December 31, 2019, when it is considered to be an unconfirmed retirement. This potential retirement of Brandon Unit 5 s approximately 95 MW of capacity is not expected to have an impact on reliability as other resources are expected to come into service at that time. Manitoba Hydro has up to 925 MW of firm and/or expected capacity exports in the winter, up to 625 MW of firm and/or expected capacity imports in the winter, and up to 1,525 MW of firm and/or expected capacity exports in the summer. There are associated firm transmission reservations over the 10 year assessment period. Manitoba Hydro does not have any capacity imports during the summer. Manitoba Hydro does not have any capacity transactions beyond the contract terms. Included in the firm exports are up to 475 MW of firm contracts in relation to the 630 MW Keeyask generating station, which is anticipated to come into service around Manitoba does not have a legislated renewable mandate, such as an RPS, and no legislation is currently anticipated. 90

100 Chapter 6: Regional Overview MRO-SaskPower Saskatchewan is a province of Canada and comprises a geographic area of 651,900 square kilometers (251,700 square miles) with approximately 1.1 million people. Peak demand is experienced in the winter. The Saskatchewan Power Corporation (SaskPower) is the Planning Coordinator and Reliability Coordinator for the province of Saskatchewan and is the principal supplier of electricity in the province. SaskPower is a Provincial Crown Corporation and under provincial legislation is responsible for the reliability oversight of the Saskatchewan bulk electric system and its interconnections. Summary of Methods and Assumptions Reference Margin Level No change in Reference Margin Level since the 2015 LTRA. The reserve margin for SaskPower s generation system must not fall below 11 percent of adjusted net demand. This percentage represents the amount of excess generation SaskPower will have after serving the highest projected load during the peak month. Load Forecast Method Coincident, 50/50 forecast Peak Season Winter Planning Considerations for Wind Resources 10 percent of nameplate (summer); 20 percent of nameplate (winter) Planning Considerations for Solar Resources No utility-scale solar resources Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 3,724 3,761 3,852 3,874 3,901 3,959 4,007 4,048 4,111 4,159 Demand Response Net Internal Demand 3,639 3,676 3,767 3,789 3,816 3,874 3,922 3,963 4,026 4,074 Resources (MW) Anticipated 4,303 4,364 4,700 4,910 4,872 5,076 5,101 5,072 5,112 5,152 Prospective 4,303 4,364 4,700 4,910 4,950 5,156 5,276 5,247 5,287 5,327 Reserve Margins (%) Anticipated 18.24% 18.72% 24.76% 29.58% 27.67% 31.01% 30.04% 27.97% 26.96% 26.46% Prospective 18.24% 18.72% 24.76% 29.58% 29.71% 33.07% 34.50% 32.38% 31.31% 30.75% Reference Margin Level 11.00% 11.00% 11.00% 11.00% 11.00% 11.00% 11.00% 11.00% 11.00% 11.00% Shortfall (MW) Anticipated Prospective

101 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: Saskatchewan is a winter-peaking area (December). The Saskatchewan Power Corporation (SaskPower) is the principal supplier of electricity in the province of Saskatchewan, Canada. SaskPower is a provincial Crown Corporation, which under the provincial legislation, is responsible for the reliability oversight of the Saskatchewan bulk electric system and is obligated to serve its domestic load. Modeling Characteristics: SaskPower utilized the Multi-Area Reliability Simulation (MARS) program for the purpose of this study. This reliability study is based on the DSM adjusted 50/50 load forecast. Results Trending: Since the 2014 Probabilistic Assessment, the reported forecast reserve margin for year 2018 has gone down slightly from 20.6 percent to 17.8 percent mainly due to a change in the expansion sequence. As expected, EUE and LOLH have increased when compared to analysis completed in Probabilistic vs. Deterministic Reserve Margin Results: Most of the data is consistent with the LTRA except the energy forecast and the expansion sequence, which has been updated to reflect the most recent projections. Base Case Study The major contribution to the 2018 LOLH and EUE is in the month of October (around 60 percent). There are maintenances scheduled to the largest coal and large natural gas units in that month. Most of the maintenance is scheduled during off-peak periods and can be rescheduled to mitigate short-term reliability issues. In the year of 2020, the LOLH and EUE are highest in January due to higher load. Sensitivity Case Study A similar monthly trend is observed in the Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) Sensitivity Case. As compared to the Base Case, the reserve margin has decreased from 17.8 percent to 15.4 percent and from 25.6 percent to 20.7 percent for year 2018 and 2020, respectively. The effect of higher load growth is evident on the reliability metrics. EUE is almost doubled from the Base Case in both study years. EUE reported for Sensitivity Case is MWh/yr and 147 MWh/yr for the year 2018 and 2020, respectively. 92

102 Chapter 6: Regional Overview Demand, Resources, and Planning Reserve Margins Saskatchewan plans to meet projected load requirements with anticipated resources throughout the assessment period. Saskatchewan s Anticipated Reserve Margin exceeds the 11 percent Reference Margin Level for the assessment period. Saskatchewan experiences peak demand in the winter. The average annual growth rate for total internal demand is 1.4 percent during the assessment period, which is slightly lower than last year s forecast (1.89 percent). The slight decrease is mainly caused by the deferral of oil pipeline projects. The growth is expected to be generally spread throughout the province. Saskatchewan is planning for a seven percent yearly average growth of energy efficiency and conservation programs, and DR programs are projected to remain unchanged. In Saskatchewan, projected unit retirements for the assessment period include 174 MW of natural gas facilities, 11 MW of wind facilities, and two-139 MW coal facilities. There were 228 MW (nameplate) added in the Assessment area since the 2015 LTRA. Throughout the assessment period, a total capacity of 2633 MW (nameplate) of Tier 1 resources is projected to come on-line. This total consists of 660 MW of natural gas, 1607 MW of wind, 120 MW of solar, 76 MW of biomass resources, 100 MW of flare gas resources, 20 MW of geothermal, and 50 MW of hydro resources. 93

103 Chapter 6: Regional Overview For capacity transfers, Saskatchewan has a firm import contract for 25 MW until the spring of Saskatchewan also has a firm import of 100 MW from July 2020 until the end of the assessment period. There are no anticipated firm exports for the assessment period. Saskatchewan only imports and exports based on economics. Import also increases supply mix diversification for Saskatchewan. Transmission Outlook and System Enhancements Saskatchewan plans to invest in transmission infrastructure over the assessment period in order to maintain and enhance reliability. The related projects are dependent on load growth and include the construction of 918 km of new 138 kv and 230 kv transmission line. Saskatchewan is also adding a static VAR system in the South-Central Region of the province to help with voltage control in the area by the end of Long-Term Reliability Issues It is not expected that extreme weather events will impact long-term reliability in Saskatchewan; however, operation of the Saskatchewan system would be performed on a best-effort basis under extreme weather events. Demand would be offset by planning reserves and external markets. If necessary, operational measures include DR, interruptible load contracts, public appeals, and rotating outages. Typically, a significant amount of unit maintenance (partial and total unit outage) is planned for the shoulder periods in Saskatchewan. If short-term reliability issues are identified during a shoulder period, unit maintenance will be rescheduled. Saskatchewan does not expect any long-term reliability impacts resulting from fuel supply and/or transportation constraints. Fuel disruptions are minimized as much as possible by system design practices and Saskatchewan s diverse energy mix of resources. Coal resources have firm contracts, are mine-to-mouth, and stockpiles are maintained at each facility in the event that mine operations are unable to meet the required demand of the generating facility. Natural gas resources have firm transportation contracts with large natural gas storage facilities located within the province capable of supporting those contractual requirements. Hydro facilities/reservoirs are fully controlled by Saskatchewan, and long term hydrological conditions are monitored. Essential Reliability Services The Saskatchewan net demand ramping trend for historical years (2013, 2014, and 2015) shows a gradual increase by approximately five percent each year. Contribution of the VERs to the increase in ramp rate was, however, minimum as there has been no significant increase in VERs in the historical years in Saskatchewan. Projected 2016 ramp rates also show approximately a five percent increase from the historical years. Projected 2020 ramp rates show an increase by approximately 70 percent from the 2016 and historical year ramp rates. Saskatchewan system inertia did not change significantly for the 2015, 2016, and 2018 years. There has not been a significant change in installed generation capacity in these years. 94

104 Chapter 6: Regional Overview NPCC-Maritimes The Maritimes assessment area is a winter-peaking NPCC subregion that contains two Balancing Authorities. It is comprised of the Canadian provinces of New Brunswick, Nova Scotia, and Prince Edward Island; as well as the northern portion of Maine, which is radially connected to the New Brunswick power system. The area covers 58,000 square miles, with a total population of 1.9 million people. Summary of Methods and Assumptions Reference Margin Level 20 percent Load Forecast Method Coincident; 50/50 forecast Peak Season Winter Planning Considerations for Wind Resources Estimated capacity is derived from a combination of mandated capacity factors and reliability impacts. Planning Considerations for Solar Resources N/A Footprint Changes A conceptual tie line to the Canadian province of Newfoundland and Labrador could potentially impact the Maritimes footprint. Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 5,584 5,608 5,627 5,623 5,622 5,608 5,580 5,552 5,509 5,518 Demand Response Net Internal Demand 5,312 5,336 5,355 5,351 5,350 5,336 5,309 5,281 5,238 5,248 Resources (MW) Anticipated 6,716 6,585 6,661 6,661 6,661 6,661 6,661 6,661 6,661 6,655 Prospective 6,735 6,621 6,735 6,735 6,735 6,735 6,735 6,735 6,735 6,724 Reserve Margins (%) Anticipated 26.42% 23.40% 24.37% 24.47% 24.49% 24.82% 25.47% 26.13% 27.16% 26.81% Prospective 26.79% 24.08% 25.77% 25.87% 25.89% 26.21% 26.88% 27.54% 28.59% 28.13% Reference Margin Level 20.00% 20.00% 20.00% 20.00% 20.00% 20.00% 20.00% 20.00% 20.00% 20.00% Shortfall (MW) Anticipated Prospective

105 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: The Maritimes Area is a winter peaking area with separate jurisdictions and regulators in New Brunswick, Nova Scotia, Prince Edward Island (PEI), and Northern Maine. The GE MARS model, developed by the NPCC CP-8 Working Group, was used for the following: demand uncertainty modeling, scheduled outages and deratings, forced outages and deratings, assistance over interconnections with neighboring Planning Coordinator Areas, transmission transfer capabilities, and capacity and/or load relief from available operating procedures, as prescribed by the NPCC resource adequacy criterion (i.e., NPCC Regional Reliability Reference Directory No. 1 Design and Operation of the Bulk Power System). 84 Results Trending: The previous study, the NERC RAS Probabilistic Assessment NPCC Region, 85 estimated an annual LOLH = hours per year and a corresponding EUE equal to 0.0 (ppm) for the year The 2018 forecast 50/50 peak demand forecast is 262 MW greater in this assessment than reported in the previous assessment. This reflects increases in electric heating loads that were not quite offset by declines in industrial loads and demand shifting programs. Forecast capacity resources increased by 81 MW in the 2016 Probabilistic Assessment as compared to the previous assessment. No material difference in estimated LOLH and EUE is observed between the two assessments. Increased capacity resources, coupled with reliance on operating procedures and tie benefits, contribute to this result. Probabilistic vs. Deterministic Reserve Margin Results: The Maritimes Area employs a reserve criterion of 20 percent of firm load. To relate the Maritimes Area reserve criterion of 20 percent to the NPCC resource adequacy criterion, LOLE was evaluated with the Maritimes Area firm load scaled so that the reserve was equal to 20 percent. The results showed that a Maritimes Area reserve of 20 percent corresponds to an LOLE of approximately days per year. Modeling: Assumptions used in this probabilistic assessment are consistent with those used in the NPCC 2016 Long Range Adequacy Overview and described in the 2016 NERC RAS Probabilistic Assessment NPCC Region NPCC Regional Reliability Reference Directory #1: Design and Operation of the Bulk Power System; September NERC RAS Probabilistic Assessment: NPCC Region; March NPCC Library - Resource Adequacy 96

106 Chapter 6: Regional Overview Base Case Study No significant LOLH is observed. EUE is in 2018 and negligible in Anticipated Reserve Margins are well above 20 percent in both years. The greatest contribution to the LOLH and EUE occur during the peak (winter) monthly period. Sensitivity Case Study LOLH is also not significant in this case, the EUE values are negligible: 0.03 and MWh for 2018 and 2020, respectively. Anticipated Reserve Margins remain above 20 percent in 2018 and near 20 percent in Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

107 Chapter 6: Regional Overview Overview The Maritimes Area is comprised of four subareas: New Brunswick (NB), Nova Scotia (NS), Prince Edward Island (PEI), and Northern Maine (NM), where a 20 percent reserve margin target is used. This reserve level correlates closely to the amount of reserve necessary to meet the 0.1 days per year LOLE criterion required by the Northeast Power Coordinating Council (NPCC). The close correlation is confirmed annually during NPCC resource adequacy reviews. Resources in the Maritimes Area are subject to capacity deratings that account for variances in seasons in the case of variable generation and the likelihood of the resource being available if called upon. If derated resources result in inadequate margins looking forward in time, then new resources must be acquired. The aggregated load growth for the four combined subareas of the Maritimes Area is practically flat for both the summer and winter seasonal peak load periods. They have an average growth rates of 0.2 percent per year in summer and a decline of 0.13 percent per year in winter. Load growth for the southeastern corner of the NB subarea (including the much smaller PEI subarea) is not specifically identified in the load projections, but NB has historically outpaced growth in the rest the Maritimes Area. NB Power has added under-voltage load shedding equipment at two new sites and applied temperature compensation to restricting lines to increase transfer capabilities. In addition, demand side management (DSM) programs aimed at reducing and shifting peak demands and potential imports to NB from NS could reduce transmission loading in the southeastern NB area. These imports may begin after the completion of the highvoltage direct current (HVdc) interconnection to the Canadian province of Newfoundland and Labrador. No other reinforcements are planned at this time. During the 10-year LTRA assessment period in the Maritimes Area, annual amounts for summer peak demand reductions associated with energy efficiency programs rose from 7 MW to 92 MW, and the annual amounts for winter peak demand reductions rose from 43 MW to 555 MW. Most of this amount is related to intensive demand shifting programs in New Brunswick that will focus mainly on reducing and/or shifting water and space heater demands during peak load periods. This is done by using smart metering technology to control their consumption patterns. Interruptible loads are the only specifically forecasted DR programs in the Maritimes Area. During the assessment period, no significant changes from previously reported amounts for interruptible loads are expected to occur. Nova Scotia is projecting three installed capacity additions by January 2020: 49 MW of wind, 3.6 MW of solar, and 23 MW of biomass/biogas. In addition, transmission restrictions on a 45 MW formerly energy-only biomass generator in Nova Scotia are being removed before January 2018, and the generator s capacity will then be counted as firm capacity. Also, an unconfirmed retirement of a 153 MW coal-fired unit in Nova Scotia is expected in mid This capacity will be offset by an expected firm purchase of hydro capacity over the new HVdc link with the Canadian province of Newfoundland and Labrador. The retirement of the coal unit will be correspondingly delayed should a delay occur in the introduction of energy from the new hydro capacity. Renewable energy standards (RESs) have led to the development of substantially more wind generation capacity than any other renewable generation type. Reduced frequency response is associated with wind generation and may, with increasing levels in the future, require displacement of wind generation with conventional generation during light load periods. With the significant amount of large scale wind energy currently being balanced on the NB system, the next phase of renewable energy development in NB will focus on smaller scale projects with a particular emphasis on nonintermittent forms of generation, such as wood-based biomass. In NS, the Maritimes Link project will provide renewable hydro resources that may otherwise have been provided by intermittent resources and would have further reduced frequency response capability. 98

108 Chapter 6: Regional Overview With respect to capacity deratings for renewable variable generation, NB derates wind capacity using a calculated year-round equivalent capacity of 20 percent for the Maritimes Area. NS and PEI derate wind capacity to 12 percent and 15 percent of nameplate based on calculated year-round equivalent capacities for their respective sub areas. NM derates wind to 26 percent and 46 percent of nameplate based on summer and winter seasonal capacity factors, respectively. There are no trends developing for either imports or exports; however, a long-term import contract starting in mid-2020 is expected to offset the unconfirmed retirement of a coal unit in Nova Scotia. This offset is expected to be with replacement hydro capacity over the HVdc cable link (currently under construction to Newfoundland and Labrador. The capacity of the import and the unit retirement are comparable and are planned to coincide in timing, thus overall resource adequacy is unaffected by these changes. Transmission development in the Maritimes Area during the assessment period includes installation of a 345 kv breaker in series with an existing breaker at NB s Point Lepreau terminal in the spring of This will mitigate contingencies and reduce import restrictions from New England. During the winter of 2016/17, the installation of two undersea 138 kv cable connections, each with a capacity of 200 MVA and a length of 9 miles, will be completed. This will increase capacity and improve the ability to withstand transmission contingencies in the area between NB and PEI. A 475 MW HVdc undersea cable link (Maritime Link) between Newfoundland and Labrador and NS will be installed by early The Maritime Link could also potentially provide a source for imports from NS into NB that would reduce transmission loading in the southeastern NB area. In addition, during the fall of 2018, a second 345/138 kv transformer will be added in parallel with an existing unit at the Keswick terminal in NB to mitigate the effects of transformer contingencies at the terminal. No delays are expected for these projects. No modifications have been made to the assessment area s planning assumptions or methods in response to extreme weather events. The hydro-electric power supply system in the Maritimes Area, with a capacity of approximately 1330 MW, is predominantly run-of-the-river as opposed to storage based. Large quantities of energy cannot be held in reserve to stave off drought conditions. If such conditions occur, the hydro system would still be used to follow load in the area and respond to sudden short-term capacity requirements. Thermal units would be used to keep the small storage capability of the hydro systems available only for load following and/or peak supply. The Maritimes Area is not overly reliant on wind capacity to meet resource adequacy requirements. The lack of wind during peaks (or very high wind speeds and/or icing conditions that would cause wind farms to suddenly shut down) should not affect the dependability of supply to the area. This is because ample spinning reserve is available to cover the loss of the largest base loaded generator in the area. The latter situation is mitigated further by wide geographic dispersal of wind resources across the area. The Maritimes Area has a diversified mix of capacity resources fueled by nuclear, oil, coal, natural gas, dual fuel oil/natural gas, hydro, wind (derated), and biomass with no one type feeding more than about 26 percent of the total capacity in the area. There is not a high degree of reliance upon any one type or source of fuel. The Maritimes Area does not anticipate fuel disruptions that pose significant challenges to resource adequacy in the area during the assessment period. This resource diversification also provides flexibility to respond to any future environmental issues, such as potential restrictions to greenhouse gas emissions. The Maritimes Area has begun tracking the ramp rate variability trend, but does not yet have enough historical years of data for the area as a whole to identify any trends. Given the essentially flat load growth and small degree of anticipated VER installations, little change in either ramp rates or the area s resource mix is expected to occur for the duration of the LTRA assessment period. The maximum net demand ramping variability (1 hour up, 1 hour down, 3 hours up, and 3 hours down values) for two historical years of 2014 and 2015 and a future year of 2020 were calculated along with the percentage contributions of VERs versus the loads. The majority of the maximums occurred during the late fall shoulder and winter peak seasons. The Maritimes Area is a winter peaking area. Five 99

109 Chapter 6: Regional Overview minute interval samples were used for these calculations. The values for 2020 were scaled up from the actuals used for The following table outlines the results of the Maritimes area review. NDRV stands for net demand ramping variability, and VER stands for variable energy resource. Year Variable NDRV (MW) Date VER (MW) Load (MW) VER/load (%) hour UP 637 Dec. 8, , hour DOWN -617 Nov. 17, , hours UP 1262 Dec. 5, , hours DOWN Nov. 17, , hour UP 606 Feb. 2, , hour DOWN -416 Dec. 17, , hours UP 1172 Oct. 28, , hours DOWN -953 Dec. 17, , hour UP 629 Feb. 2, , hour DOWN -434 Dec. 17, , hours UP 1219 Oct. 28, , hours DOWN -993 Dec. 17, ,

110 Chapter 6: Regional Overview NPCC-New England ISO New England (ISO-NE) Inc. is a regional transmission organization that serves Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont. ISO-NE is responsible for the reliable day-to-day operation of New England s bulk power generation and transmission system; ISO-NE also administers the area s wholesale electricity markets and manages the comprehensive planning of the regional BPS. The New England regional electric power system serves approximately 14.5 million people over 68,000 square miles. Summary of Methods and Assumptions Reference Margin Level The installed capacity requirement (ICR) results in a Reference Margin Level of percent in 2017, declining to percent in 2018 and expected to be percent for the remainder of the assessment period. Load Forecast Method Coincident; normal weather (50/50) Peak Season Summer Planning Considerations for Wind Resources A value of 5 percent of the total nameplate for on-shore and 20 percent of nameplate for off-shore resources Planning Considerations for Solar Resources A value of 38 percent of nameplate in 2017, decreasing annually to 34 percent in 2026 Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 26,698 26,765 26,783 26,789 26,816 26,870 26,942 27,026 27,122 27,218 Demand Response Net Internal Demand 25,857 26,168 26,405 26,411 26,438 26,492 26,564 26,648 26,744 26,841 Resources (MW) Anticipated 31,112 32,529 32,617 31,226 31,330 31,336 31,342 31,347 31,353 31,353 Prospective 31,313 32,935 33,803 32,412 32,516 32,522 32,528 32,533 32,539 32,539 Reserve Margins (%) Anticipated 20.32% 24.31% 23.52% 18.23% 18.50% 18.28% 17.99% 17.63% 17.23% 16.81% Prospective 21.10% 25.86% 28.02% 22.72% 22.99% 22.76% 22.45% 22.08% 21.67% 21.23% Reference Margin Level 16.74% 16.55% 15.93% 15.93% 15.93% 15.93% 15.93% 15.93% 15.93% 15.93% Shortfall (MW) Anticipated Prospective

111 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview The New England Area is a summer peaking area comprised of New Hampshire, Rhode Island, and Vermont. The GE MARS model developed by the NPCC CP-8 Working Group was used for the following: modeling demand uncertainty, scheduled outages and deratings, forced outages and deratings, assistance over interconnections with neighboring Planning Coordinator Areas, transmission transfer capabilities, and capacity and/or load relief from available operating procedures. This is as prescribed by the NPCC resource adequacy criterion (ref: NPCC Regional Reliability Reference Directory No. 1 Design and Operation of the Bulk Power System). 87 Results trending: The previous study (NERC RAS Long-Term Reliability Assessment NPCC Region) 88 estimated an annual LOLH equal to hours per year and a corresponding EUE equal to MWh for the year The Forecast 50/50 peak demand for 2018 was lower than reported in the previous study with higher estimated forecast planning and forecast operable reserve margins. As a result, both the LOLH and the EUE have improved for Probabilistic vs. Deterministic Reserve Margin Results: New England s reference reserve margin is determined based on the NPCC resource adequacy criterion; this results in a reference reserve margin level of 16.6 percent in 2018, and 15.9 percent for Modeling: Assumptions used in this probabilistic assessment are consistent with those used in NPCC 2016 Long Range Adequacy Overview, and are consistent with those described in the 2016 NERC RAS Probabilistic Assessment NPCC Region NPCC Regional Reliability Reference Directory #1: Design and Operation of the Bulk Power System; September NPCC RAS Probabilistic Assessment; March NPCC Library - Resource Adequacy 102

112 Chapter 6: Regional Overview Base Case Study In 2018, LOLH is h/year and EUE is 65.2 MWh while in 2020 those values are h/year and MWh, respectively. The increases are consistent with a decline in reserve margins. The metrics are primarily driven by the results in July and August. Sensitivity Case Study LOLH and EUE increase exponentially with the decline in reserve margins. LOLH is and h/year for 2018 and 2020, respectively. EUE is and MWh for those two years. As it was the case in the Base Case, July and August have the biggest share of the annual metrics. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

113 Chapter 6: Regional Overview Overview ISO New England s (ISO-NE) Reference Margin Level is based on the capacity needed to meet the Northeast Power Coordinating Council (NPCC) 1-in-10-year LOLE resource planning reliability criterion. The amount of capacity needed, referred to as the installed capacity requirement (ICR), varies from year-to-year, depending on expected system conditions. The capacity needed to meet the LOLE criterion is purchased through annual forward capacity auctions three years in advance. Reconfiguration auctions occur annually prior to the commencement year to assure an opportunity to adjust capacity purchases to meet changing requirements. ISO-NE s Anticipated Reserve Margin, which ranges from a high of 24.3 percent in 2018 to a low of 16.8 percent in 2026, remains above the Reference Margin Level through the assessment period. The summer peak total internal demand (TID), which takes into account energy efficiency and conservation as well as behind-the-meter photovoltaic (PV) resources, is forecasted to increase from 26,698 MW in 2017 to 27,218 MW in This amounts to a nine-year summer TID compound annual growth rate (CAGR) of 0.21 percent, as compared to the 2015 LTRA projection of 0.48 percent. The primary reasons for the decrease in the demand forecast are updated historical data, and an increased amount of behind-the-meter PV. As behind-the-meter PV resources increase, New England could experience daily load profiles that would require different resource operating attributes to manage reserve, ramping, and regulation requirements. Both passive and active DR are procured through ISO-NE s forward capacity market (FCM). Passive DR, which consists of energy efficiency and conservation, will grow to 2,561 MW by 2019 in the FCM. For the years beyond the FCM commitment periods, ISO-NE uses an energy efficiency forecasting methodology that takes into account the potential impact of growing energy efficiency and conservation initiatives throughout the region. Energy efficiency has generally been increasing over time and is projected to continue growing throughout the study period. The amount of energy efficiency is projected to increase to over 4,000 MW by Active demand resources consist of DR and emergency generation, which can be activated with the implementation of ISO-NE Operating Procedure No. 4: Action during a Capacity Deficiency (OP-4). 90 The capacity supply obligations (CSOs) for these resources, which are obtained through ISO-NE s FCM, decrease from 556 MW in 2016 to 378 MW in A total of 319 MW of new capacity consisting primarily of PV and wind resources have been added since the 2015 LTRA. Approximately 2,900 MW of Tier 1 capacity, including over 2,700 MW of natural-gas-fired plants, will be added by The largest natural gas projects are the Footprint Combined Cycle Plant (674 MW), which is projected to be in service in 2017, and the CPV Towantic Energy Center (725 MW) as well as PSEG s Bridgeport Harbor Expansion (484 MW), both of which are to be in service in Also included in the Tier 1 category is 133 MW of on-peak wind capacity (806 MW nameplate). Tier 2 capacity additions totaling 1,012 MW include 982 MW of natural-gas-fired generation and 125 MW of nameplate wind. The amount of renewable resources in New England continues to grow. In addition to behind-the-meter PV that reduces the load forecast, there has been growth in PV participating in ISO-NE markets, with on-peak capacity increasing from 264 MW in 2017 to 329 MW in Although the amount of Existing and Tier 1 wind capacity only amounts to 229 MW on peak, there is an additional 3,400 MW of nameplate wind capacity in the ISO-NE generator interconnection queue. Over 2,100 MW of retirements are expected in New England by Brayton Point Station, which is a 1,535 MW coal and oil plant, will be retiring by June The 680 MW Pilgrim Nuclear Power Station is planned for retirement by June Even with these retirements, ISO-NE s reserve margin is not expected to fall below the 15.9 percent Reference Margin Level during the assessment period. If capacity is required to meet the regional resource 90 OP-4 is used by ISO-NE operators when resources are insufficient to meet the anticipated load plus operating reserve requirement. 104

114 Chapter 6: Regional Overview adequacy, ISO-NE will purchase the needed capacity through its forward capacity market (FCM). Currently, over 680 MW of Tier 2 capacity has a capacity supply obligation in the FCM. The major transmission project currently under development in New England is the Greater Boston project. The Greater Boston upgrades are critical to improve the ability to move power into the Greater Boston area and from northern New England to southern New England. This set of upgrades includes new and upgraded 345 kv and 115 kv lines, new autotransformers, and additional reactive support. The project is certified to be in service by June For over a decade, the region has been working on gas-related challenges and will continue to do so. New England s generation fleet is changing rapidly with the retirement of older fossil-fueled resources and their replacement by new gas-fired generators. The region s reliance on natural gas for power generation has been increasing and will likely continue to do so in the future. ISO is addressing the gas-related challenges with market rule changes and operational enhancements. Recent market rules, such as those addressing energy market offer flexibility, allow resources to more accurately reflect their variable costs in their energy offers during the operating day, which improves incentives to perform. Another new market rule has changed the timing of the day-ahead energy market to align more closely with natural gas trading deadlines. In addition, the ISO has improved coordination and information-sharing with natural gas pipeline operators, such as working with the pipelines to coordinate generator and pipeline maintenance schedules. The ISO has also developed a natural gas usage tool that estimates the remaining gas pipeline capacity, by individual pipe, for use by ISO-NE system operators to determine whether the electric sector gas demand can be accommodated. A winter fuel-reliability program has been acting as a bridge between now and 2018 when the longer-term pay-forperformance (PFP) capacity market changes go into effect. The winter reliability program addresses regional winter reliability challenges created by New England s increased reliance on natural-gas-fired generation and lack of adequate gas infrastructure. Resources participating in the program provide incremental energy inventory during the winter months to help ensure reliable system conditions. Components of the program include payment to generators for adding dual-fuel capability, securing fuel inventory, testing fuel-switching capability, compensation for any unused fuel inventory, and nonperformance charges. PFP, which starts in June 2018, will help improve reliability while ensuring resource adequacy. PFP is a two-part settlement in which a base payment is set in the forward capacity auction and a performance payment is determined during the delivery year. The performance payment may be positive or negative, depending on resource performance during a shortage condition. Over-performing resources are paid a premium through revenue transfers from under-performing resources. PFP creates an incentive for investment in generators that are either: 1) low-cost and highly reliable (nearly always operating), or 2) highly flexible and highly reliable (goes on-line quickly and reliably). PFP will encourage generators to increase unit availability by implementing dual-fuel capability, entering into firm gas-supply contracts, and investing in new fast-responding assets. By creating incentives for generators to firm up their fuel supply, PFP may indirectly provide incentives for the development of on-site oil, LNG fuel storage, or expanded gas pipeline infrastructure. In summary, New England has adequate capacity resources to meet the NERC Reference Margin Level throughout the 2016 LTRA study period. ISO New England has been faced with gas-related challenges for more than a decade. These challenges will remain as additional non-gas-fired resources retire and are replaced by gas-fired generation. ISO New England expects and continues to make operational and market enhancements to address these challenges. ISO New England is cognizant of possible operational issues that a high penetration of intermittent resources may pose in the future. The region has conducted and will continue to conduct studies to identify means to address these future operational issues. Furthermore, New England has a robust transmission planning process. Existing and planned transmission upgrades will ensure regional system reliability. 105

115 Chapter 6: Regional Overview NPCC-New York The New York Independent System Operator (NYISO) is the only BA within the state of New York (NYBA). NYISO is a single-state ISO that was formed as the successor to the New York Power Pool a consortium of the eight IOUs in NYISO manages the New York State transmission grid, which encompasses approximately 11,000 miles of transmission lines over 47,000 square miles and serving the electric needs of 19.5 million New Yorkers. New York experienced its all-time peak load of 33,956 MW in the summer of Summary of Methods and Assumptions Reference Margin Level The New York State Reliability Council (NYSRC) installed reserve margin (IRM) of 17.5 percent applies to the period May 2016 to April New York s IRM is set annually. Load Forecast Method The New York Balancing Authority (NYBA) forecast is based upon an econometric forecast of annual energy and seasonal peak demands. The New York State Reliability Council (NYSRC) has adjustments for energy efficiency and DERs, including behind-the-meter solar PV. Peak Season The seasonal peak demands (summer and winter) are based upon annual energy and seasonal load factors. The forecast load factors are based upon recent historic data and then trended for the future. Planning Considerations for Wind Resources The expected on-peak capacity for wind resources is 14 percent. Planning Considerations for Solar Resources On peak resources for solar PV include a number of factors, such as inverter sizing and efficiency, the impact of cloud cover and other atmospheric conditions that attenuate solar irradiance, and the seasonal and diurnal variations in solar irradiance. Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 33,363 33,404 33,477 33,501 33,555 33,650 33,748 33,833 33,926 34,056 Demand Response 1,248 1,248 1,248 1,248 1,248 1,248 1,248 1,248 1,248 1,248 Net Internal Demand 32,115 32,156 32,229 32,253 32,307 32,402 32,500 32,585 32,678 32,808 Resources (MW) Anticipated 39,613 40,056 40,065 40,727 40,727 40,727 40,727 40,727 40,727 40,727 Prospective 40,382 40,923 42,805 43,474 43,474 43,474 43,474 43,474 43,474 43,474 Reserve Margins (%) Anticipated 23.35% 24.57% 24.31% 26.27% 26.06% 25.69% 25.31% 24.99% 24.63% 24.14% Prospective 25.74% 27.26% 32.81% 34.79% 34.56% 34.17% 33.76% 33.42% 33.04% 32.51% Reference Margin Level 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Shortfall (MW) Anticipated Prospective

116 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: The New York Area is a summer peaking area. The GE MARS model developed by the NPCC CP-8 Work Group was used for the following: modeling demand uncertainty, scheduled outages and deratings, forced outages and deratings, assistance over interconnections with neighboring Planning Coordinator Areas, transmission transfer capabilities, and capacity and/or load relief from available operating procedures, as prescribed by the NPCC resource adequacy criterion. 91 Results Trending: The previous study, the NERC RAS Long-Term Reliability Assessment NPCC Region 92 estimated an annual LOLH = hours per year and a corresponding EUE equal to 9.3 MWh for the year The Forecast 50/50 Peak Demand for 2018 was lower than reported in the previous study, but with higher estimated forecast planning and forecast operable reserve margins. As a result, both the LOLH and the EUE have improved for Probabilistic vs. Deterministic Reserve Margin Results: The New York IRM of 17.5 percent applies to the period May 2016 to April New York s IRM is set annually. New York does not have a future reference reserve margin beyond the current capability period thus the NERC reference reserve margin is used. Modeling: Assumptions used in this probabilistic assessment are consistent with those used in the NPCC 2016 Long Range Adequacy Overview, and it is described in the 2016 NERC RAS Probabilistic Assessment NPCC Region. 94 All New York published reports and probabilistic studies report reserve margins based on the full ICAP value of all resources. 91 NPCC Regional Reliability Reference Directory #1: Design and Operation of the Bulk Power System; September NPCC NERC RAS Probabilistic Assessment; March New York State Reliability Council: New York Control Area Installed Capacity Requirements Reports 94 NPCC Library - Resource Adequacy 107

117 Chapter 6: Regional Overview Base Case Study LOLH for 2018 and 2020 are (hours per year) with EUE values of and (MWh). The EUEs are negligible. Results are similarly driven by a comparable planning reserve margin in both years. The summer months (June August) have the greatest contribution to these metrics. Sensitivity Case Study LOLH values are and for 2018 and 2020, respectively. EUE results are 2.8 and 7.6 MWh for those same two years. The monthly contribution is similar to that observed in the Base Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated * Prospective * NERC Reference ProbA Forecast Planning ** ProbA Forecast Operable ** Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) * NERC LTRA reserve margin calculations are based on a 14 percent wind unit peak capacity factor. ** Wind units modeled in the probabilistic assessment as hourly load modifiers are based on 2013 production data, and ProbA capacity resource interconnection service values are used for reserve margin calculations. 108

118 Chapter 6: Regional Overview Planning Reserve Margins The annual IRM for the New York Balancing Area (NYBA) is calculated through a technical study conducted by the New York State Reliability Council (NYSRC) in accordance with NERC Reliability Standards, the Northeast Power Coordinating Council (NPCC) reliability criteria, and the NYSRC Reliability Rules. For the capability year, the NYSRC approved an IRM requirement of 17.5 percent. New York does not have a future reference reserve margin beyond the current capability period, ending April The New York IRM assumed that the ~1,455 MW of wind capacity will likely operate at a 14 percent capacity factor during the summer peak period. This assumed capacity factor is based on an analysis of actual hourly wind generation data collected for wind facilities in New York State during the June through August 2013 period between 2:00 and 5:00 p.m. This test period was chosen because it covers the time during which virtually all of the annual NYCA LOLE occurrences are. For the calculation of New York IRM, wind generators are modeled as hourly load modifiers. In the probabilistic determination of the New York IRM, the output of each unit varies between 0 MW and the capacity resource interconnection service value based on 2013 production data. All generator values for the IRM requirement calculation are based on generator installed capability values as reported in the 2016 NERC LTRA and the current Load and Capacity Data Report issued by the New York Independent System Operator, Inc. (NYISO). For reporting purposes, the capacity values provided for New York existing and planned resource facilities are consistent on its dependable maximum net capability (DMNC). In circumstances where the ability to deliver power to the grid is restricted, the value of the resource is limited to its capacity resource interconnection service (CRIS) value. The source of DMNC ratings for existing facilities are seasonal tests required by procedures in the NYISO Installed Capacity Manual and are documented in the New York Gold Book. 95 Demand The baseline energy forecast for the years is expected to decline at an average rate of percent per year, while the baseline summer peak demand forecast for the years is expected to grow at annual average rate of 0.21 percent. The lower forecasted growth in energy usage for years is largely attributable to an increase in the impacts of energy efficiency initiatives and the growth of distributed behind-themeter energy resources. Demand-Side Management There were no significant changes to NYISO s DR programs since last year s LTRA. The 2016 forecast includes peak demand impacts of 1) energy efficiency initiatives in the amount of 1,859 MW, 2) solar PV in the amount of 747 MW, and 3) distributed generation in the amount of 356 MW. These are cumulative impacts expected by the year DR enrollments are currently trending at approximately 4.3 percent of the NYBA s peak load, and there is no indication that there will be a significant increase in enrollment in the near future. The NYISO does not anticipate significant long-term reliability impacts from a modest increase in the DR enrollments from the current enrollment levels. Generation Since the 2015 LTRA, there were no new resource additions installed in the NYBA. Tier 1 resource additions total 775 MW and are expected to be in service for Summer Tier 2 resources total 4,140 MW and are at various stages in the NYISO interconnection process. However, the NYBA had MW of summer capacity deactivate, and they have another 1,734.8 MW of capacity scheduled to deactivate over the assessment period. The NYISO 95 NYISO: 2016 Load & Capacity Data 109

119 Chapter 6: Regional Overview did not identify any near-term reliability needs resulting from these deactivations. Other than these deactivations, no large generators are expected to be unavailable over the assessment period. In response to an April 2016 order from the Federal Energy Regulatory Commission (FERC), NYISO is further developing a tariff process to be filed in September 2016 to address reliability needs that arise from generator deactivations and the planning process for identifying solutions; this includes the potential need for a reliabilitymust-run agreement to keep the deactivating generator in service until permanent solutions can be provided. The forecast for distributed behind-the-meter generation for the summer peak demand is 313 MW in 2021 and 356 MW in DERs are expected to increase in the future with behind-the-meter solar growing at the fastest rate. Projected additions are based upon current rates of growth in each NYBA zone, together with an expectation of how future state incentives for distributed generation will affect installation of these resources. Changing Resource Mix New York State had a renewable portfolio standard (RPS) that has been supplanted by other state programs, including a large-scale renewables program under the New York State Clean Energy Fund (CEF). The RPS program purchased renewable energy credits from seventy active projects that represented 2,152 MW. These programs produced more than 5,000 GWH in 2015 and more than 90 percent of this renewable generation was produced by wind resources. Currently, the New York State Energy Plan calls for 50 percent of energy generation to come from renewable energy sources by The New York State Public Service Commission (NYSPSC) is currently developing a clean energy standard that is intended to achieve the goals of the New York State Energy Plan and to preserve the financial viability of the existing nuclear generators in New York. The New York State Department of Public Service staff estimates that the proposal calls for 75,000 GWH of annual renewable energy production by The NYSPSC has not yet finalized the specific types of renewable generation that will be included under the CES. One of New York State s initiatives, the NY-Sun Incentive Program (NY-Sun), is designed to have 3,000 MW of installed solar PV capacity on the system by the end of In April 2014, following two successful years of solar PV installations, a commitment of nearly $1 billion was made to NY-Sun for further installation of solar PV. In response to the increasing amount of VERs and New York State s initiatives, the NYISO studied a number of specific grid operation needs potentially affected by the increasing penetration of intermittent solar PV and wind resources. The study found, among other things, that 1) the bulk power system can reliably manage over the fiveminute time horizon the increase in net load variability associated with the solar PV and wind penetration levels up to 4,500 MW wind and 9,000 MW solar PV, 2) the large-scale implementation of behind-the-meter solar PV will impact the NYISO s load profile and associated system operations, and 3) the lack of frequency and voltage ride-through requirements for solar PV facilities in New York could worsen system contingencies when solar PV deactivates in response to frequency and voltage excursions. Likewise, wind resources in New York are increasing and now total approximately four percent of the generation fleet by fuel source. A 2010 NYISO wind generation study 96 examined the impact of adding up to 8,000 MW of wind resources and it indicated that, above 3,500 MW of wind penetration, regulation requirements are projected to increase at the rate of 25 MW for every 1,000 MW increase in wind generation. NYISO has not changed the methods that it uses to determine the on-peak capacity values for wind, solar, and hydro. Hourly unit output data for wind, run of river hydro, and solar units are collected for the summer peak hours (i.e., 2:00 p.m. 5:00 p.m.) from June 1 through August 31. The on-peak capacity for these resources is determined using an assumed capability for each resource class; this is based upon unit historic operating data and engineering judgment. For reserve margin calculations, NYISO uses the full on-peak capability of the units, 96 NYISO: Growing Wind: Final Report of the NYISO 2010 Wind Generation Study; September

120 Chapter 6: Regional Overview which represents the aggregate capacity for each resource class (i.e., wind, solar, and hydro). The expected onpeak capacity factors for wind, solar, and hydro are 14 percent, 56.5 percent, and 54 percent respectively. Capacity Transfers NYISO has three classifications of capacity transfers: The first includes grandfathered contracts and external capacity resource interconnection service (CRIS) rights. These total 2170 MW and cover the entire 2016 LTRA assessment period. The second class is unforced deliverability rights (UDRs). These are rights to deliver capacity over controllable tie lines. For the NYBA, the total UDR capability is 1,965 MW across the four controllable tie lines. The owners of the UDRs notify NYISO each year of the amount of capacity that will be delivered. UDR election levels are treated by NYISO as confidential information. Any transfer capability not utilized is available to provide emergency assistance in both the NYISO s planning studies and operationally, if the need arises. The third classification of capacity transfers is import rights. For 2016, import rights totaled 530 MW and are available month to month on a first-come, first-served basis in the capacity auctions. Capacity transactions modeled in the NYISO s assessments have met the capacity resource requirements, as defined in the NYISO s tariffs. Both NYISO and its respective neighboring assessment areas have agreed upon the terms of the capacity transaction including the MW value, the duration, the contract path, the source of capacity, and the capacity rating of the resource. Transmission and System Enhancements The NYBA has three major transmission projects located in central New York, downstate New York, and New York City, all placed into service in June These include Marcy-South Series Compensation and Fraser-Coopers Corners 345 kv line reconductoring, construction of a second Rock Tavern-Ramapo 345 kv line, and Phase I (i.e., cable separation) of upgrading underground transmission circuits from Staten Island to the rest of New York City (collectively, Transmission Owner Transmission Solutions or TOTS ). Approved by the NYSPSC as part of New York s Energy Highway initiative, the TOTS projects are expected to increase transfer capability into southeastern New York by 450 MW and mitigate against potential reliability needs if the Indian Point Energy Center were to become unavailable. In the 2014 Reliability Needs Assessment (RNA), NYISO identified thermal violations under N-1-1 post-contingency conditions (applying more stringent NPCC criteria) that would limit transmission in the Rochester and Syracuse areas. For the Rochester area, the overloads are on 345/115kV transformers that supply the Rochester area upon loss of other 345/115kV transformers in the same area; the Syracuse area overloads on 115kV facilities upon loss of parallel lines. These violations are anticipated to be resolved with permanent solutions identified in the most recent Transmission Owner local transmission plans, scheduled to be completed by Summer 2017 in the Rochester area and the end of 2017 in the Syracuse area. In the interim, the local transmission owners will implement local operating procedures, if required, to prevent overloads, including the potential for limited load shedding in the Rochester and Syracuse areas; voltage-constrained transfer limits are evaluated and determined by NYISO for all major interfaces within New York. BPS transmission security is maintained by limiting power transfers according to the determined voltage-constrained transfer limits. Local nonbulk voltage performance is evaluated by the local Transmission Owner and addressed through the Local transmission planning process. Transmission security of the NYBA BPS is maintained by limiting power transfers according to the determined transfer limits, including voltage-constrained transfer limits. New York has three interfaces that were found to be voltage limited, and NYISO maintains voltage limits in these constrained areas by limiting power transfers to 111

121 Chapter 6: Regional Overview mitigate dynamic and static reactive power issues. Based on the foregoing, NYISO does not expect to use undervoltage load shedding schemes. Depending on assumed system conditions, the Central East interface is limited at certain times due to dynamic instability. As part of the annual NYISO Area Transmission Review (ATR), the flows on the evaluated interfaces were tested at a value of at least 10 percent above the more restrictive of the emergency thermal or voltage transfer limits in accordance with NYISO Transmission Planning Guideline # The 2014 intermediate ATR performed dynamic stability simulations for NERC contingencies that were expected to produce the more severe system results or impacts based on examination of actual system events and assessment of changes to the planned system. BPS transmission security is maintained by limiting power transfers according to the determined stability limits. The New York Balancing Authority will also have additions and removals of special protection systems (SPSs) since the last LTRA. Generation rejection SPSs are being retired at the Niagara hydro facility and the St. Lawrence Moses hydro facility, which will become effective upon completion of the NPCC approval process. Both facilities have added power system stabilizers to their units and a study showed that thermal limits, voltage, and stability would be maintained for the contingencies at those facilities. Additionally, an SPS is being added to mitigate subsynchronous resonance issues when the new series compensated lines in the TOTS projects are placed in service in Summer The SPS will detect certain outage conditions and signal to bypass the series compensation. Long-Term Reliability Issues NYISO continues to plan for extreme weather and has not made any modifications to its planning assumptions or methods for such events. NYISO continues to conduct its reliability studies using the 50/50 load forecast as the base assumption and account for weather events with a load forecast uncertainty (LFU). Additionally, NYISO, in conjunction with its stakeholders, is exploring market rule changes to help assure fuel availability during cold weather conditions. Improvements will be considered in reporting seasonal fuel inventories and daily replenishment schedules. NYISO will work with New York State regulatory agencies to develop a formal process to identify reliability needs that would be mitigated by generator requests for certain waivers. NYISO only conducts dynamic stability studies for the off-peak periods and, in doing so, identified no concerns. Generators in the fleet use the off-peak period to schedule and perform their routine maintenance in preparation for the summer and winter peak seasons. NYISO monitors and approves maintenance schedules to maintain system reliability and can cancel scheduled maintenance if system conditions warrant it. As a part of NYISO s 2014 Comprehensive Reliability Plan 98 and the 2015 Power Trends report 99, NYISO identified several risk factors to maintaining reliability in New York. These factors include the following: Changes to System Performance: The aging generation infrastructure may lead to more frequent and longer outages as well as increasing costs, which may drive more retirements. Since 2000, more than 11,000 MW of generation has been added while more than 6,000 MW are no longer active. Of the current generation fleet, 8.5 GW are produced by generators that are more than 50 years old. This figure is expected to double by 2025 in the absence of new generation being built to replace aging assets. Accelerated or unplanned retirements can present challenges to system reliability. Furthermore, the preliminary results of the 2016 RNA show that if the remaining nuclear generation units on the system were to deactivate, NYISO would see immediate resource adequacy needs. 97 NYISO Transmission Expansion and Interconnection Manual; October NYISO 2014 Comprehensive Reliability Plan; March NYISO Power Trends: Rightsizing the Grid;

122 Chapter 6: Regional Overview Changes to System Load: The potential for higher-than-forecasted system loads under the probability level could expose the system to potential reliability issues, including greater levels of load shedding in the interim operating procedures in some localized areas of the state. Changes to System Resources: Expected capacity resources (new or upgrades) within the study do not materialize, additional generating units become unavailable or retired beyond those already identified, or capacity resources could decide to offer into other markets and, therefore, not be available to New York. Natural Gas Coordination: Coordinating with New York s reliance on natural gas as the primary fuel for electric generation, NYISO is performing four ongoing studies and efforts focused on 1) improving communication and coordination between the sectors; 2) addressing market structure enhancements, such as the closing time of the natural gas markets; 3) providing for back up fuel (primarily distillate oil) assurance to generation; and 4) addressing the electric system reliability impact of the sudden catastrophic loss of gas. Federal and State Environmental Regulations: The five regulatory programs with the largest reliability risk potential include: 1) facility specific operational limitations, 2) the Cross State Air Pollution Rule (CSAPR) cap and trade program for NOx and SO2, 3) the Mercury and Air Toxics Standards (MATS) for hazardous air pollutants from new and existing coal and oil-fired units, 4) the CPP, which is the proposed EPA greenhouse gas standards for existing sources, and 5) the revised Ozone National Ambient Air Quality Standard (NAAQS). Furthermore, the New York State CES seeks to reduce the state s carbon dioxide emissions by 40 percent; the CES seeks to do this through increasing the amount of renewable energy generation in New York State to 50 percent of total energy production by The New York State Department of Public Service staff estimates that to meet the CES s goals by 2030, energy from renewables would need to increase by 33,700 GWh from the current levels. Based on historical demonstrated capacity factors, NYISO estimates that this increase will require the development of approximately 25,000 MWs of solar capacity, approximately 15,000 MWs of wind capacity, or approximately 4,000 MWs of hydroelectric capacity. NYISO continues to study the impacts of the relative capability of intermittent resources to reliably supply power demands and fulfill IRM requirements. Essential Reliability Services NYISO is conducting a study on the reliability impacts of the EPA CPP. A portion of this study will examine changes to essential reliability service (ERS) metrics as the resource mix changes, and as the inertia and kinetic energy, as presented in the NERC ERS Task Force Measures Framework Report. This assessment will be performed by compiling the inertia and MVA rating for all NYBA generators and tabulating in each hour, based on future model run year output, which generators are operating and the aggregate system kinetic energy that corresponds. 113

123 Chapter 6: Regional Overview NPCC-Ontario The Independent Electricity System Operator (IESO) is the balancing authority for the province of Ontario. The province of Ontario covers more than one million square kilometers (415,000 square miles) and has a population of more than 13 million people. Ontario is interconnected electrically with Québec, MRO-Manitoba, states in MISO (Minnesota and Michigan), and NPCC-New York. Summary of Methods and Assumptions Reference Margin Level The IESO-established reserve margin requirement is applied as the Reference Margin Level. 100 Load Forecast Method Coincident; normal weather (50/50) Peak Season Summer Planning Considerations for Wind Resources Modeled, based on historic performance and historic weather data Planning Considerations for Solar Resources Modeled, based on historic weather data Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 22,680 22,519 22,357 22,192 22,479 22,255 22,190 22,194 22,326 22,265 Demand Response ,007 1,210 1,210 Net Internal Demand 22,000 21,878 21,756 21,591 21,878 21,654 21,386 21,188 21,116 21,056 Resources (MW) Anticipated 26,822 26,431 27,216 27,478 26,235 25,872 24,957 25,773 23,819 24,646 Prospective 26,822 26,431 27,216 27,478 26,290 26,000 25,085 25,901 23,947 24,837 Reserve Margins (%) Anticipated 21.92% 20.81% 25.10% 27.27% 19.92% 19.48% 16.70% 21.64% 12.80% 17.05% Prospective 21.92% 20.81% 25.10% 27.27% 20.17% 20.07% 17.30% 22.25% 13.41% 17.96% Reference Margin Level 18.13% 17.31% 17.13% 17.67% 17.00% 17.00% 18.00% 18.00% 16.00% 16.00% Shortfall (MW) Anticipated Prospective Ontario IESO, for its own assessments, treats demand response as a resource instead of as a load modifier. As a consequence, the net internal demand, planning reserve margins, and target reserve margin numbers differ in IESO reports when compared to NERC reports. The Ontario reports would show lower reserve margins. 114

124 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: The Ontario Area is a summer peaking area. The GE MARS model developed by the NPCC CP-8 Working Group was used for the following: modeling demand uncertainty, scheduled outages and deratings, forced outages and deratings, assistance over interconnections with neighboring Planning Coordinator Areas, transmission transfer capabilities, and capacity and/or load relief from available operating procedures, as prescribed by the NPCC resource adequacy criterion (ref: NPCC Regional Reliability Reference Directory No. 1 Design and Operation of the Bulk Power System). 101 Results Trending: The previous study, NERC RAS Long-Term Reliability Assessment NPCC Region, 102 estimated an annual LOLH = 0.0 hours per year and a corresponding EUE equal to 0.0 (ppm) for the year The 2018 forecast 50/50 peak demand forecast is 218 MW greater in this assessment than reported in the previous assessment. This reflects the interplay of economic expansion, population growth, increasing penetration of electrically powered devices, conservation programs, increasing embedded generation output, and energy price changes that act to reduce the amount of grid-supplied electricity needed. There is no change in the estimated LOLH and EUE between the two assessments mainly due to the contributions of various DR programs, operating procedures, and tie benefits. Probabilistic vs. Deterministic Reserve Margin Results: The Ontario IESO, in its own assessments, treats DR as a resource instead of a load modifier. As a consequence, reserve margin calculations are lower in IESO reports when compared to NERC assessments. Modeling: Assumptions used in this probabilistic assessment are consistent with those used in NPCC 2016 Long Range Adequacy Overview and described in the 2016 NERC RAS Probabilistic Assessment NPCC Region NPCC: Regional Reliability Reference Directory #1 102 NERC RAS Probabilistic Assessment: NPCC Region; March 31, NPCC Library - Resource Adequacy 115

125 Chapter 6: Regional Overview Base Case Study There was no significant LOLH or EUE observed for the Base Case study for either 2018 or Anticipated Reserve Margins are above percent and percent in 2018 and 2020, respectively. Sensitivity Case Study LOLH values are not significant in this case, and the EUE are negligible:.004 and.074 MWh for 2018 and 2020, respectively. Anticipated Reserve Margins remain above the Base Case reference reserve margin in both years. The greatest contribution to EUE occurs during the peak (summer) monthly period. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

126 Chapter 6: Regional Overview Supply-Demand Balance and Resource Adequacy Ontario has enough confirmed planned resources (Tier 1) to meet its Reference Margin Levels in all years except for 2023 and The analysis shows that the earliest need for additional resources may arise in 2023, and that need is expected to be less than 1 GW. Ontario possesses a range of options to address these needs, including market-based mechanisms and capacity imports. Over the next ten years, Ontario expects grid-connected electricity demand to decline slightly, both in terms of annual energy and summer peak. While modest economic and population growth is expected, increases in demand are expected to be offset by three key factors: Growth in distributed generation, driven in large part by government renewable capacity targets and feed-in tariff programs 13 TWh of annual conservation savings (incremental to today s demand levels), driven by updates to codes and standards, conservation incentives, and energy efficiency programs The continuing success of peak-reduction incentive programs that are already in-place, such as the Industrial Conservation Initiative and time-of-use rates Over the past ten years, Ontario has invested heavily in electricity infrastructure to enable the phase-out of coalfired generation and to reduce the carbon footprint of Ontario s electricity supply. The next ten years will also be marked by further change as the system continues its transformation. Retirements Pickering nuclear station, with an installed capacity of about 3 GW or 8.6 percent of Ontario s current supply, is expected to be decommissioned between 2022 and Nuclear Refurbishments 8.5 GW of nuclear supply at Darlington and Bruce nuclear plants is expected to undergo mid-life refurbishment between 2016 and Much of this occurs during the assessment period, with up to 4 nuclear unit s off-line during a refurbishment outage simultaneously during the peak refurbishment year. The development of the refurbishment programs was supported by Ontario s past experience and the plan will be implemented in a way that minimizes risk. Capacity Additions Ontario expects to add 3.5 GW of grid-connected generating capacity over the assessment period, of which just over 1 GW is natural gas, and the balance is renewable resources such as wind and solar. Demand Response IESO continues to transition the procurement of DR from capacity-based DR (CBDR) programs to an annual DR auction. This is a transparent and cost-effective way to select the most competitive providers of DRs while ensuring that all providers are held to the same performance obligations. The first DR competitive auction was held in December 2015, where nearly 400 MW were procured. Ontario currently has approximately 550 MW of CBDR and DR Auction capacity under contract, a similar level to that in last year s LTRA. At minimum, this level of capacity will be maintained through subsequent auctions with additional capacity-based DR expected to be acquired between 2021 and 2025, consistent with government targets, to a total of 1,200 MW by Ontario currently has over 1.1 GW of DR capability. It is anticipated that DR capacity will reach 1.8 GW by the end of the assessment period, consistent with government targets. 117

127 Chapter 6: Regional Overview Distributed Generation Over the assessment period, a further 1.1 GW of variable generation is expected to be added to the distribution system. This is in addition to the 3.7 GW of variable generation currently connected at the distribution level. Transmission Outlook and System Enhancements Transmission planning to address changes to the supply mix and ensure reliability throughout the province is ongoing. Two major system enhancement projects are underway: 1) a new 230 kv double-circuit East-West Tie line in Northwestern Ontario and 2) a new 500 to 230 kv transformer station (TS), Clarington TS, in the Eastern portion of the Greater Toronto Area. The expected in-service date for the new East West Tie line is 2020, and the Clarington transformer station is scheduled to be in service in Planning studies are being finalized to manage the loading on the transmission lines between Trafalgar TS and Richview TS and the 500/230 kv transformers at Claireville TS and Trafalgar TS, which are forecasted to be exceeded by Planning options have been assessed and are expected to include the installation of 500/230 kv autotransformers at the existing Milton Switching Station (SS) with eight 230 kv circuit terminations and 12 km of new double-circuit line sections connecting the new Milton TS to Hurontario SS. Southern Ontario experiences high voltages during light load periods, and with the planned shutdown of Pickering GS and the removal of its reactive absorption capability, the situation is expected to persist. Planning work for the installation of new voltage control devices is being finalized. Long-Term Reliability Issues With the growth in distributed generation capacity, demand forecasting has become increasingly more complex. Traditionally, demand was mainly a function of weather conditions, economic cycles, and population growth. With multiple new factors influencing demand, such as increased distribution-connected variable generation and increased consumer price-responsiveness, determining the causality of demand changes has become increasingly nuanced. The introduction of variable generation (e.g., solar and wind) and the removal of flexible generation (e.g., coal), combined with lower demand and limitations in operational flexibility of gas and hydro resources, have added new challenges to maintaining a reliable system. The results of a recent operability assessment indicated that there is a system need for enhanced flexibility to balance supply and demand, more regulation, and additional grid voltage control. It is important that the supply mix remain robust in meeting industry planning standards, flexible to meet the ever-changing demands of system operations, and balanced in managing inherent risks, such as fuel security and critical infrastructure needs. To that end, the IESO has launched an initiative to augment resource flexibility and issued a request for information for additional regulation service in June IESO has an energy storage pilot program underway to test the capability of storage technologies to provide grid services as well. Activities are also underway with transmitters to plan and install additional dynamic and static voltage control devices to help with voltage control. Increasing amounts of variable generation, coupled with relatively flat demand levels, have contributed to a rise in surplus baseload generation (SBG) in Ontario. Over the next few years, more variable generation is expected, but the effects on SBG will be tempered by the impact of the planned nuclear refurbishments and retirements. The IESO has mechanisms in place to manage SBG, including economic exports, wind and solar dispatch, and nuclear maneuvers or shutdowns. 118

128 Chapter 6: Regional Overview NPCC- Québec The Québec assessment area (Province of Québec) is a winter-peaking NPCC subregion that covers 595,391 square miles with a population of eight million. Québec is one of the four NERC Interconnections in North America, with ties to Ontario, New York, New England, and the Maritimes, consisting either of HVdc ties, radial generation, or load to and from neighboring systems. Summary of Methods and Assumptions Reference Margin Level Reference margin levels are drawn from the Québec Area 2015 Interim Review of Resource Adequacy, which was approved by NPCC s Reliability Coordinating Committee in December Load Forecast Method Coincident; normal weather (50/50) Peak Season Winter Planning Considerations for Wind Resources On-peak contribution is approximately 30 percent of the total Planning Considerations for Solar Resources N/A Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 38,150 38,521 38,875 39,130 39,415 39,689 39,939 40,167 40,388 40,625 Demand Response 2,168 2,238 2,318 2,318 2,318 2,318 2,318 2,318 2,318 2,318 Net Internal Demand 35,982 36,283 36,557 36,812 37,097 37,371 37,621 37,849 38,070 38,307 Resources (MW) Anticipated 41,217 41,847 42,348 42,746 42,746 42,746 42,746 42,746 42,746 42,746 Prospective 42,317 42,947 43,448 43,846 43,846 43,846 43,846 43,846 43,846 43,846 Reserve Margins (%) Anticipated 14.55% 15.34% 15.84% 16.12% 15.23% 14.38% 13.62% 12.94% 12.28% 11.59% Prospective 17.61% 18.37% 18.85% 19.11% 18.19% 17.33% 16.55% 15.85% 15.17% 14.46% Reference Margin Level 12.20% 12.70% 12.70% 12.70% 12.70% 12.70% 12.70% 12.70% 12.70% 12.70% Shortfall (MW) Anticipated Prospective

129 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: Québec is a winter peaking area. The GE MARS model developed by the NPCC CP-8 Working Group was used for the following: modeling demand uncertainty, scheduled outages and deratings, forced outages and deratings, assistance over interconnections with neighboring Planning Coordinator Areas, transmission transfer capabilities, and capacity and/or load relief from available operating procedures, as prescribed by the NPCC resource adequacy criterion (ref: NPCC Regional Reliability Reference Directory No. 1 Design and Operation of the Bulk Power System). 104 Results Trending: The previous study, NERC RAS Long-Term Reliability Assessment NPCC Region, 105 estimated an annual LOLH = 0.0 hours per year and a corresponding EUE equal to 0.0 for the year The forecast 50/50 peak demand for 2018 was lower than reported in the previous study with a slightly higher estimated forecast planning and forecast operable reserve margins. As a result, there is no change in the estimated LOLH and EUE in this year s study. Probabilistic vs. Deterministic Reserve Margin Results: Québec s Reference Reserve Margin is determined based on the NPCC resource adequacy criterion; results indicate a Reference Reserve Margin of 12.7 percent. 106 Modeling: Assumptions used in this probabilistic assessment are consistent with those used in NPCC 2016 Long Range Adequacy Overview and described in the 2016 NERC RAS Probabilistic Assessment NPCC Region NPCC: Regional Reliability Reference Directory #1 105 NERC RAS Probabilistic Assessment: NPCC Region; March 31, NPCC 2015 Québec Balancing Authority Area Interim Review of Resource Adequacy; December NPCC Library - Resource Adequacy 120

130 Chapter 6: Regional Overview Base Case Study No LOLH or EUE was estimated for 2018 or The Anticipated Reserve Margins are above the Reference Reserve Margins for 2018 and 2020, respectively. Sensitivity Case Study No LOLH or EUE was estimated for 2018 or The Anticipated Reserve Margins are near the Reference Reserve Margins. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

131 Chapter 6: Regional Overview Demand, Resources, and Planning Reserve Margins The Prospective Reserve Margin remains above the Reference Margin Level for all seasons and years during the assessment period. Under the Prospective Reserve Margin, a total of 1,100 MW of expected capacity imports are planned by the Québec Area. These purchases have not yet been backed by firm long-term contracts; however, on a yearly basis, the Québec Area proceeds with short-term capacity purchases (UCAP) if needed to meet its capacity requirements. The Québec area demand forecast average annual growth is 0.7 percent during the 10-year period; this is similar to last year s forecast. Total internal demand is calculated for the Québec area as a single entity and the area s peak demand forecast is coincident. DR programs in the Québec Area specifically designed for peak-load reduction during winter operating periods are mainly interruptible load programs (for large industrial customers), totaling 1,748 MW for the winter period. The total on-peak DR for the winter period is projected to be 2,318 MW. In 2015, the generating station La Romaine-1 was integrated for a total of 270 MW of new added hydro capacity. Work is under way on the La Romaine-3 (395 MW) development which will be fully operational in Some preparatory work has also begun on the La Romaine-4 (245 MW) development, which will be fully operational in The integration of small hydro units also account for 83 MW of new capacity during the assessment period. For other renewable resources, about 350 MW (105 MW on-peak value) of wind capacity and 5 MW of biomass have been added to the system since the beginning of Additionally, 663 MW (199 MW on-peak value) of wind capacity and 128 MW of biomass are expected to be in service by The Québec Area will support firm capacity sales totalling 750 MW during the winter peak period, declining to 145 MW for the winter period and after. Transmission Outlook and System Enhancements This section reviews several major transmission projects currently underway. The Romaine River Hydro Complex Integration Construction of the Romaine River Hydro Complex project is presently underway. Its total capacity will be 1,550 MW. Romaine-2 (640 MW) has been commissioned in December 2014, and Romaine-1 (270 MW) in December Romaine-3 (395 MW) and Romaine-4 (245 MW) will be integrated in at Montagnais 735/315-kV substation. The Québec area is reiterating its commitment to sustainable development by focusing on renewable energy at the Romaine complex, which will help meet current needs without jeopardizing the energy supply of future generations. Main system upgrades for this project have required construction of the new Aux Outardes 735-kV switching station, located between existing Micoua and Manicouagan substations. Two 735-kV lines have been redirected into the new station, and one new 735-kV line (5 km or 3 miles) has been built between Aux Outardes and Micoua substations. This upgrade has been commissioned in Summer The Chamouchouane Montréal 735-kV Line Planning studies have shown the need to reinforce the transmission system with a new 735-kV line in the near future in order to meet the Reliability Standards. The line will extend from the Chamouchouane substation on the eastern James Bay subsystem to a new substation (Judith Jasmin) in Montréal (about 400 km or 250 miles). The new 735kV substation is required to fulfill two objectives: 1) providing a new source of electricity supply on the north shore of Montreal and 2) connecting the new 735kV line from Chamouchouane to the Montreal metropolitan loop. This project will reduce transfers on other parallel lines on the Southern 735-kV Interface, thus 122

132 Chapter 6: Regional Overview optimizing operation flexibility and reducing losses. The line is scheduled for the winter peak period. Public information meetings have been held and construction phase has begun. Upcoming Regional Projects Other regional substations and/or line projects are in the planning/permitting stages. There are about a dozen regional transmission projects in the Montréal and Québec City areas. There are another dozen projects in other areas with in service dates from 2016 to 2020, consisting mostly of 315/25-kV and 230/25-kV distribution substations to replace 120-kV and 69-kV infrastructures. Long-Term Reliability Issues While technical developments in recent years have contributed to create a more reliable system, sustainable system reliability may be challenged by emerging issues, such as potential operational issues due to the changing resource mix. In Québec area, wind generation capacity has increased by 2,500 MW over the five last years, but the area s total installed capacity is still mainly composed of large reservoir hydro complexes (more than 90 percent). These complexes can react quickly to adjust their generation output and meet the sharp changes in electricity net demand. The forecasted change to resource mix is not expected to have any influence on the ramp rate trends or any other reliability issue. 123

133 Chapter 6: Regional Overview PJM PJM Interconnection is a regional transmission organization (RTO) that coordinates the movement of wholesale electricity in all or parts of Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia, and the District of Columbia. PJM companies serve 61 million people and cover 243,417 square miles. PJM is a Balancing Authority, Planning Coordinator, Transmission Planner, Resource Planner, Interchange Authority, Transmission Operator, Transmission Service Provider, and Reliability Coordinator. Summary of Methods and Assumptions Reference Margin Level The PJM RTO reserve requirement is applied as the Reference Margin Level for this assessment. Load Forecast Method PJM has significantly revised its load forecast model. The treatment of weather has been restructured to provide more variable load response across a wide range of conditions. Three variables (cooling, heating, and other) were added to account for trends in equipment/appliance saturation and efficiency. Distributed solar generation is now reflected in the historical load data used to estimate the models. This is done with a separately-derived solar forecast that is used to adjust load forecasts. Peak Season Summer Planning Considerations for Wind Resources Initially, 13 percent of nameplate replaced with historic information tracked over the peak period Planning Considerations for Solar Resources Initially, 38 percent of nameplate replaced with historic information tracked over the peak period Footprint Changes The East Kentucky Power Cooperative (EKPC), which integrated into the PJM RTO on June 1, 2013, is now part of PJM s load and generation data. Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 154, , , , , , , , , ,891 Demand Response 8,883 8,977 9,035 3,416 3,424 3,436 3,450 3,478 3,499 3,524 Net Internal Demand 145, , , , , , , , , ,367 Resources (MW) Anticipated 190, , , , , , , , , ,178 Prospective 194, , , , , , , , , ,353 Reserve Margins (%) Anticipated 31.11% 33.50% 33.79% 28.48% 28.09% 27.58% 26.78% 25.98% 25.23% 24.51% Prospective 33.95% 37.88% 46.01% 50.38% 52.54% 51.96% 51.33% 50.37% 49.48% 48.61% Reference Margin Level 16.50% 16.50% 16.50% 16.50% 16.50% 16.50% 16.50% 16.50% 16.50% 16.50% Shortfall (MW) Anticipated Prospective NERC 2016 Long-Term Reliability Assessment DRAFT October

134 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: The probabilistic assessment was carried out in GE-MARS using Monte Carlo simulation. Internal and external load shapes were from year 2002 (Summer) and 2004 (Winter) adjusted to match monthly and annual peak forecast values from the 2016 PJM load forecast. Data on individual unit performance is from the period PJM was divided in five subareas interconnected using a transportation/pipeline approach. External areas were modeled using a detailed representation (NPCC) and at planned reserve margin (MISO, TVA, VACAR). Modeling: Load forecast uncertainty was modeled on a monthly basis using a normal distribution discretized in seven steps 108. Demand side management (DSM) was modeled as an emergency operating procedure as most of the DSM in PJM is emergency DSM. Intermittent generators were modeled as a regular resource at their respective capacity values (average capacity value for wind is 13 percent while for solar is 38 percent). Firm exports/imports were explicitly modeled while the limits on the transportation/pipeline interfaces were calculated based on a First Contingency Total Transfer Capability (FCTTC) analysis. Results trending: The 2018 LOLH and EUE in the 2016 ProbA are smaller than the corresponding values reported in the 2014 ProbA: 2018 LOLH in 2016 ProbA = hrs/year vs LOLH in 2014 ProbA = hrs/year 2018 EUE in 2016 ProbA = MWh/year vs EUE in 2014 ProbA = MWh/year This difference can be explained by the larger planning and operable reserves for 2018 in the 2016 ProbA compared to those in the 2014 ProbA. The increase in 2018 reserves is due to a reduction in net internal demand and an increase in forecast capacity resources. In particular, the increase in forecast capacity resources is due to the fact that, by the time the 2014 ProbA was run, none of the 2018 capacity market auctions had been cleared. In contrast, the forecast capacity resources for 2018 considered in the 2016 ProbA include capacity secured via capacity market auctions. Probabilistic vs. Deterministic Reserve Margin Results: For Summer 2018 and Summer 2020, the probabilistic reserve margin in slightly lower than the deterministic value due to 2,500 MW of on-peak capacity derates as a result of above average summer ambient conditions. 108 PJM: Load Forecasting and Analysis; June

135 Base Case Study LOLH is zero for both 2018 and 2020 due to large forecast planning reserve margins (significantly above the reference value of 16.5 percent). EUE is virtually zero (though technically nonzero) for both 2018 and The only month that contributes a discernible amount of EUE in both years is April due to planned maintenance and large load uncertainty for some of the areas within PJM. Chapter 6: Regional Overview Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) Sensitivity Case Study LOLH is still zero for For year 2020, LOLH exhibits a very mild uptick (i.e., hours per year) during April due to a large amount of planned maintenance and large load uncertainty for some of the areas within PJM. EUE is slightly higher than under the Base Case for both 2018 and 2020 but still very close to zero. Months that contribute to the EUE in the Sensitivity Case are April (due to the reasons mentioned above explaining the LOLH uptick in 2020) and July (where the PJM annual peak occurs). 126

136 Chapter 6: Regional Overview Summary The PJM RTO reserve requirement, as calculated by PJM, is 16.4 percent for the 2016/2017 planning period, which runs from June 1, 2016 through May 31, The PJM RTO reserve requirement is 0.8 percentage points higher this year compared to the 2015/2016. About three-eighths of this increase can be attributed to changes in the PJM load model: shorter historical time period, greater energy efficiency and distributed generation, and more granular weather monitoring. Another three-eighths can be attributed to worse performance by the generation units while the remaining quarter is due to reduced emergency imports from the world (i.e., outside the PJM area). Since the modeling of the PJM peak is nearer to the world peak, there is a lack of diversity with the world peak. A 16.5 percent PJM RTO reserve requirement is applicable for the rest of the assessment period. PJM RTO will have an adequate Anticipated Reserve Margin though the entire assessment period. The prospective margin is also adequate for the entire assessment period. Since the 2015 report, PJM has significantly revised its load forecast model. The treatment of weather has been restructured to provide more variable load response to weather across a wide range of conditions. Three variables (cooling, heating, and other) were added to account for trends in equipment/appliance saturation and efficiency. Distributed solar generation is now reflected in the historical load data used to estimate the models with a separately-derived solar forecast used to adjust load forecasts. 109 The winter load forecast has smaller changes compared to the summer load forecast. This is due to two impacts acting against one another to minimize changes to previous forecasts: 1) the result of shortening the historical period that PJM uses to produce weather scenarios and 2) this lowered the resulting winter forecast. The other impact is related to refinements to the weather specification that addressed the previous model s tendency to understate the elasticity of load to weather at peak conditions. The PJM Capacity Performance initiative (a PJM program to incentivize better generator performance) starts to show up in future DR accounting since DR is considered capacity in PJM. This program actually decreases DR by more than half since performance is required the entire year and some DR programs that include air conditioning reductions cannot reduce air conditioning load that is not there. A PJM committee is investigating a seasonal aspect to capacity that may influence the amount of DR accepted by PJM in the future. PJM has begun to track residential PV installations through the PJM Generation Attribute Tracking System (GATS) since there is an effect on the PJM load forecast. Estimates (of the effect) for 2016 are 574 MW, and this increases 109 Detailed information on the development of the distributed solar generation forecast can be found on the PJM website. 127

137 Chapter 6: Regional Overview to 1,523 MW in This is less than one percent of the PJM forecasted peak load, so these energy sources will have little effect on adequacy or reliability in PJM. PJM Environmental Information Services (EIS) operates the GATS. EIS is a wholly owned subsidiary of PJM Technologies, Inc., which is a subsidiary of PJM Interconnection. The functional design of the GATS has been developed through considerable deliberation of a stakeholder group that included representatives from various state agencies (public utility commissions, environmental protection offices, energy offices, and consumer advocates), market participants, environmental advocates, and PJM staff. The design of the GATS is an unbundled certificates-based tracking system. This means that the attributes or characteristics of the generation are separated from the megawatt hour (MWh) of Energy and recorded onto a certificate after the MWh of energy is produced. Variable resources are only partially counted for PJM resource adequacy studies. Both wind and solar initially utilize class average capacity factors: 13 percent for wind and 38 percent for solar. Performance over the peak period is tracked and the class average capacity factor is supplanted with historic information. After three years of operation, only historic performance over the peak period is used to determine the individual unit's capacity factor. Biomass and hydro are counted at 100 percent of reported existing-certain resources because these resources are typically only fully utilized over the peak period of the day. Some run-of-the-river hydro capacity has always been reported as a lower value than total plant nameplate in PJM due to the full capability of the plant not typically being available. PJM has 6,748 MW of firm imports and 1,395 MW of firm exports (resulting in a net firm import of 5,353 MW) scheduled for the planning period (June 1, 2016 to May 31, 2017). Firm imports drop to 5,364 MW, 1,413 MW of firm exports with a net firm import of 3,951 MW in planning period. PJM has 4,126 MW of firm imports, 1,395 MW of firm exports with a next firm import of 2,731 MW scheduled for the planning period. The same imports and exports as the planning period are expected for the remaining years of the assessment. PJM has recently experienced below average winter temperatures. PJM s winter peak reliability analysis indicates that the transmission system is capable of delivering the system generating capacity at winter peak. PJM has experienced some thermal overload problems during light load conditions with relatively high wind generator output. PJM s light load reliability analysis 110 ensures that the transmission system is capable of delivering the system generating capacity during light load conditions. The system generating capability modeling assumption for this analysis is that the generation modeled reflects generation by fuel class that historically operates during the light load demand level like high wind output. Interchange levels for the various PJM zones will reflect a statistical average of typical previous years interchange values for off-peak hours. Load level, interchange, and generation dispatch for non-pjm areas that impact PJM facilities are based on statistical averages for previous off-peak periods. The flowgates ultimately used in the lightload reliability analysis are determined by running all contingencies maintained by PJM planning. These are also determined through monitoring all PJM market monitored facilities and BPS facilities. The contingencies used for light load reliability analysis will include NERC TPL P1, P2, P4, P5, and P7. NERC TPL P0, normal system conditions will also be studied. There has been a steady retirement of coal resources that are being replaced by combined cycle natural-gaspowered resources. No difference has been seen in net demand ramping variability related to this change in resource mix within PJM. 110 PJM Analysis of Light Load Historical Data and Light Load Reliability Criterion; July 9,

138 Chapter 6: Regional Overview SERC SERC is a summer-peaking assessment area that covers approximately 308,900 square miles and serves a population estimated at 39.4 million. SERC is divided into three assessment areas: SERC-E, SERC-N, and SERC-SE. The SERC Region includes 11 BAs: Alcoa Power Generating, Inc. Yadkin Division (Yadkin), Associated Electric Cooperative, Inc. (AECI), Duke Energy Carolinas and Duke Energy Progress (Duke), Electric Energy, Inc. (EEI), LG&E and KU Services Company (as agent for Louisville Gas and Electric (LG&E) and Kentucky Utilities (KU)), PowerSouth Energy Cooperative (PowerSouth), South Carolina Electric & Gas Company (SCE&G), South Carolina Public Service Authority (Santee Cooper, SCPSA), Southern Company Services, Inc. (Southern), and Tennessee Valley Authority (TVA). Summary of Methods and Assumptions Reference Margin Level Entities within the SERC footprint adhere to stateset targets that vary throughout the footprint. For this assessment, NERC applies a 15 percent Reference Margin Level for all SERC subregions. Load Forecast Method Noncoincident; normal weather (50/50) Peak Season Summer Planning Considerations for Wind Resources As reported by individual Generator Owners Planning Considerations for Solar Resources As reported by individual Generator Owners Footprint Changes None to report SERC-East Assessment Area Footprint SERC-North Assessment Area Footprint SERC-Southeast Assessment Area Footprint SERC-East Peak Season Demand, Resources, and Reserve Margins NERC 2016 Long-Term Reliability Assessment DRAFT October

139 Chapter 6: Regional Overview SERC-East Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 43,213 43,999 44,672 45,440 46,126 46,774 47,449 48,053 48,668 49,309 Demand Response Net Internal Demand 42,558 43,336 44,005 44,771 45,454 46,100 46,773 47,374 47,987 48,625 Resources (MW) Anticipated 51,175 51,139 51,149 51,958 54,798 56,629 56,629 57,746 57,746 58,863 Prospective 51,722 51,686 51,696 52,505 55,345 57,176 57,176 58,293 58,293 59,410 Reserve Margins (%) Anticipated 20.25% 18.01% 16.23% 16.05% 20.56% 22.84% 21.07% 21.89% 20.34% 21.06% Prospective 21.53% 19.27% 17.48% 17.27% 21.76% 24.03% 22.24% 23.05% 21.48% 22.18% Reference Margin Level 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: SERC utilizes an 8,760 hourly load, generation, and transmission simulation model that consists of three internal NERC assessment areas (SERC-E, SERC-N, and SERC-SE) and seven connected external areas (10 total external areas). First Contingency Incremental Transfer Capability (FCITC) analysis sets limits for nonfirm support amongst internal and external areas while positive and negative demand side resources represent net firm interchange schedules. Forecast assumptions for normal (50/50) coincident demand, net energy for load, and anticipated resources from the LTRA are input for the model. Then further analysis determines uncertainty parameters such as load forecast uncertainty, generator forced outage rates, etc. In addition to SERC s portion of the NERC 2016 Probabilistic Resource Assessment (PRA), SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analyses of incremental adjustments to load, resource performance, and interface limits. Also, although the U.S. Supreme Court granted a stay that halted the implementation of the Environmental Protection Agency s (EPA) CPP, a CPP scenario will address the potential resource adequacy implications of retiring coal-fired generation, further reliance on variable energy resources (VERs), and increased dependence on gas-fired generation. 130

140 Chapter 6: Regional Overview The full SERC report to be published in quarter one of 2017 will showcase enhancements to the 2016 SERC PRA, as compared to the 2014 study, as well as SERC s full sensitivity and scenario analysis. 111 Lowering demand projections in SERC-E (five percent decrease from 2014 to 2016 in the study year 2018 forecast) continue to increase Anticipated Reserve Margins and decrease the resource adequacy measures in the assessment area. Although near zero, SERC-E LOLH and EUE for both the 2018 and 2020 Base Cases contribute nearly 100 percent of the totals for the SERC assessment area footprint (SERC-E, SERC-N, and SERC-SE). However, higher excess capacity exists in the other two SERC areas. Furthermore, at anticipated load growth and reserve levels, SERC-E meets an industry standard resource planning criteria of 1-day-in-10-years LOLE in Results Trending: From the 2014 to 2016 PRA, the SERC-E LOLH decreased by approximately 97 percent (0.085 to 0.002) for the same study year This is primarily driven by the lower projected demand mentioned above as well as the 2016 modeling corrections. The SERC PRA model now includes expected firm capacity transfers and improvements to winter historical load profiles. 112 After accounting for lower demand and modeling corrections, SERC-E Base Case 2018 results remain static from Probabilistic vs. Deterministic Reserve Margin Results: For all SERC assessment areas, the probabilistic assessment (ProbA) forecast planning reserve margin is higher than the deterministic Anticipated Reserve Margin. This is due to the following differences in ProbA vs. deterministic modeling: The average probabilistic total internal demand is lower than the deterministic total internal demand due to the use of multiple load shapes with some annual peaks during non-summer months. The ProbA model optimizes scheduled maintenance so that, on average, zero maintenance occurs on peak. Controllable DR programs effective load reduction realization is higher in the ProbA model based on statistical performance rates than the deterministic value. VER performance in the ProbA model is based on a time series correlation analysis, which may be better than the expected on peak MW in the deterministic study. 111 SERC Reliability and Performance Analysis (RAPA) homepage 112 Approximations: (2014 PRA LOLH) minus (decrease load forecasts from 2014 to 2016) minus (modeling corrections) equals

141 Chapter 6: Regional Overview Base Case Study SERC-E LOLH increases to hours per year in 2018 and hours per year in EUE increases 1.4 MWh in 2018 and to 49.4 MWh in This is due to an approximate 3 percent increase in peak demand and minimal increase in anticipated resources. However, the rise of the metrics in 2020 is not concerning considering the MW size of SERC-E. Measures not modeled in the 2016 PRA such as, but not limited to, voluntary and noncontrollable DR, operating procedures to cut nonfirm schedules or maintenance, public appeals, and other mechanisms should mitigate 49.4 MWh of annual EUE within SERC-E. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) LOLH and EUE accrue relatively evenly across all months of the year in 2018; however, with increases in demand by 2020, the majority of LOLH and EUE accrues during the peak seasons of summer and winter. Actually, between 60 and 70 percent occurs during the winter months. This is contributable to a high annual 50/50 demand per unit and higher winter load forecast uncertainty due to off-normal events. A recent off-normal event was the 2014 Polar Vortex when annual peaks occurred for many entities within SERC-E during winter months. Sensitivity Case Study SERC-E entities expect a 1.44 percent compound annual growth rate (CAGR). The NERC Sensitivity Case doubles the SERC-E CAGR to 2.90 percent. In this load growth scenario, SERC-E LOLH increases to hours per year in 2018 and hours per year in EUE increases 7.6 MWh in 2018 and to MWh in SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analysis on load growth and load forecast uncertainty. This assessment will further demonstrate the influence a decline in expected energy efficiency gains and changes in other demand factors may pose to SERC-E resource adequacy and will be published quarter one of

142 Chapter 6: Regional Overview SERC-North Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 42,540 42,955 43,051 43,419 43,800 44,184 44,572 44,954 45,331 45,690 Demand Response 1,789 1,792 1,811 1,813 1,695 1,632 1,588 1,552 1,549 1,544 Net Internal Demand 40,751 41,163 41,240 41,606 42,105 42,552 42,984 43,402 43,782 44,146 Resources (MW) Anticipated 48,910 49,068 49,337 49,337 50,177 50,177 51,017 51,857 51,857 53,247 Prospective 51,265 51,351 51,620 51,620 52,460 52,460 53,300 54,140 54,140 55,530 Reserve Margins (%) Anticipated 20.02% 19.20% 19.63% 18.58% 19.17% 17.92% 18.69% 19.48% 18.44% 20.62% Prospective 25.80% 24.75% 25.17% 24.07% 24.59% 23.28% 24.00% 24.74% 23.66% 25.79% Reference Margin Level 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions 133

143 Chapter 6: Regional Overview Probabilistic Assessment Overview General Overview: SERC utilizes an 8,760 hourly load, generation, and transmission simulation model that consists of three internal NERC assessment areas (SERC-E, SERC-N, and SERC-SE) and seven connected external areas (10 total external areas). First Contingency Incremental Transfer Capability (FCITC) analysis sets limits for nonfirm support amongst internal and external areas while positive and negative demand side resources represent net firm interchange schedules. Forecast assumptions for normal (50/50) coincident demand, net energy for load, and anticipated resources from the LTRA are input for the model. Then further analysis determines uncertainty parameters such as load forecast uncertainty, generator forced outage rates, etc. In addition to SERC s portion of the NERC 2016 Probabilistic Resource Assessment (PRA), SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analyses of incremental adjustments to load, resource performance, and interface limits. Also, although the U.S. Supreme Court granted a stay that halted the implementation of the Environmental Protection Agency s (EPA) CPP, a CPP scenario will address the potential resource adequacy implications of retiring coal-fired generation, further reliance on variable energy resources (VERs), and increased dependence on gas-fired generation. The full SERC report to be published in quarter one of 2017 will showcase enhancements to the 2016 SERC PRA, as compared to the 2014 study, as well as SERC s full sensitivity and scenario analysis. 113 The demand projections in SERC-N decrease 3 percent in 2018 from the 2014 to 2016 study year. This decrease in demand forecasts continues to increase reserve margin and decrease the resource adequacy measures in the assessment area. Due to a high forecasted reserve margin of 27 percent, SERC-N experiences zero LOLH and near zero EUE for 2018 and 2020 Cases. Results Trending: From the 2014 PRA to the 2016 PRA, the SERC-N LOLH decreased in similar fashion to SERC-E, which was from to for the same study year of Again this is largely driven by the decreasing load projections. See SERC-E section for a complete synopsis of changes. Probabilistic vs. Deterministic Reserve Margin Results: For all SERC assessment areas, the probabilistic assessment (ProbA) forecast planning reserve margin is higher than the deterministic Anticipated Reserve Margin. This is due to the following differences in ProbA vs. deterministic modeling: The average probabilistic total internal demand is lower than the deterministic total internal demand due to the use of multiple load shapes with some annual peaks during non-summer months. The ProbA model optimizes scheduled maintenance so that, on average, zero maintenance occurs on peak. Controllable DR programs effective load reduction realization is higher in the ProbA model based on statistical performance rates than the deterministic value. VER performance in the ProbA model is based on a time series correlation analysis, which may be better than the expected on peak MW in the deterministic study. 113 SERC Reliability and Performance Analysis (RAPA) homepage 134

144 Base Case Study Zero LOLH and EUE. SERC-N entities expect a 0.81 percent CAGR. However, the model results for 2020 base summer yielded near zero percent growth from However, since the expected growth is below 1 percent, the resulting impact on the indices is negligible. Sensitivity Case Study The NERC Sensitivity Case doubles the SERC-N CAGR to 1.74 percent. In this load growth scenario, SERC-N Chapter 6: Regional Overview Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) LOLH and EUE increase but of minimal consequence to resource adequacy. LOLH increases to hours per year in 2018 and hours per year in EUE increases to 1.8 MWh in 2018 and to 0.8 MWh in The resulting metrics for 2020 are lower than 2018 due to gas-fired generation additions to SERC-N mid-year Subsequently, the winter months in 2020 reflect lower accrual of LOLH and EUE than in SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analysis on load growth and load forecast uncertainty. This assessment will further demonstrate the influence a decline in expected energy efficiency gains and changes in other demand factors may pose to SERC-E resource adequacy and will be published quarter one of

145 Chapter 6: Regional Overview SERC-Southeast Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 47,762 48,124 48,507 48,903 49,325 49,756 50,281 50,859 51,461 52,083 Demand Response 2,228 2,238 2,247 2,256 2,260 2,262 2,265 2,267 2,270 2,273 Net Internal Demand 45,534 45,886 46,260 46,647 47,065 47,494 48,016 48,592 49,191 49,810 Resources (MW) Anticipated 60,062 60,115 61,101 62,222 62,126 62,396 62,397 62,560 62,562 62,636 Prospective 60,596 60,659 61,645 62,765 62,669 62,939 62,940 63,103 63,105 63,179 Reserve Margins (%) Anticipated 31.91% 31.01% 32.08% 33.39% 32.00% 31.38% 29.95% 28.75% 27.18% 25.75% Prospective 33.08% 32.20% 33.26% 34.55% 33.16% 32.52% 31.08% 29.86% 28.29% 26.84% Reference Margin Level 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% 15.00% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions 136

146 Chapter 6: Regional Overview Probabilistic Assessment Overview General Overview: SERC utilizes an 8,760 hourly load, generation, and transmission simulation model that consists of three internal NERC assessment areas (SERC-E, SERC-N, and SERC-SE) and seven connected external areas (10 total external areas). First Contingency Incremental Transfer Capability (FCITC) analysis sets limits for nonfirm support amongst internal and external areas while positive and negative demand side resources represent net firm interchange schedules. Forecast assumptions for normal (50/50) coincident demand, net energy for load, and anticipated resources from the LTRA are input for the model. Then further analysis determines uncertainty parameters such as load forecast uncertainty, generator forced outage rates, etc. In addition to SERC s portion of the NERC 2016 Probabilistic Resource Assessment (PRA), SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analyses of incremental adjustments to load, resource performance, and interface limits. Also, although the U.S. Supreme Court granted a stay that halted the implementation of the Environmental Protection Agency s (EPA) CPP, a CPP scenario will address the potential resource adequacy implications of retiring coal-fired generation, further reliance on variable energy resources (VERs), and increased dependence on gas-fired generation. The full SERC report to be published in quarter one of 2017 will showcase enhancements to the 2016 SERC PRA, as compared to the 2014 study, as well as SERC s full sensitivity and scenario analysis. 114 Lowering demand projections in SERC-SE (four percent decrease from 2014 to 2016 in study year 2018 forecast) continue to increase Anticipated Reserve Margins and decrease the resource adequacy measures in the assessment area. Due to a high forecasted reserve margin of 35 percent, SERC-SE experiences zero LOLH and near zero EUE for 2018 and 2020 Base Cases. Results Trending: From the 2014 to 2016 PRA, the SERC-SE LOLH decreased in similar fashion to SERC-E and SERC-N, which was from to for the same study year of This is largely driven by the decreasing load projections. See the SERC-E section for a synopsis of corrected modeling errors from Probabilistic vs. Deterministic Reserve Margin Results: For all SERC assessment areas, the probabilistic assessment (ProbA) forecast planning reserve margin is higher than the deterministic Anticipated Reserve Margin. This is due to the following differences in ProbA vs. deterministic modeling: The average probabilistic total internal demand is lower than the deterministic total internal demand due to the use of multiple load shapes with some annual peaks during non-summer months. The ProbA model optimizes scheduled maintenance so that, on average, zero maintenance occurs on peak. Controllable DR programs effective load reduction realization is higher in the ProbA model based on statistical performance rates than the deterministic value. VER performance in the ProbA model is based on a time series correlation analysis, which may be better than the expected on peak MW in the deterministic study. 114 SERC Reliability and Performance Analysis (RAPA) homepage 137

147 Base Case Study Zero LOLH and EUE. Sensitivity Case Study SERC-SE entities expect a 1.20 percent CAGR. The NERC Sensitivity Case doubles the SERC-SE CAGR to 2.52 percent. In this load growth scenario, SERC-SE LOLH and EUE still remain zero. SERC conducts its own independent resource adequacy assessment with supplementary sensitivity analysis on load growth and load forecast uncertainty. This assessment will Chapter 6: Regional Overview Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) further demonstrate the influence a decline in expected energy efficiency gains and changes in other demand factors may pose to SERC-E resource adequacy and will be published quarter one of

148 Chapter 6: Regional Overview SERC Summary Current projections for the SERC non-rto assessment area show reserve margins in excess of 15 percent throughout the long-term planning horizon. In the near term ( ), the Region s reserve margin will range from 21.7 to 24.6 percent. The reserve margin decreases that are projected throughout subsequent long-term years ( ) are due to the uncertainty of resource additions. To maintain reserve margin levels, SERC non- RTO entities continuously plan for new capacity, acquire or request power purchase agreements, acquire additional assets, and implant new energy efficiency and DR programs. Uncommitted generating capacity in the SERC Region is not included in comparison with the NERC Reference Reserve Margin; because uncommitted capacity exists in the Region, there will continue to be additional generation above what is reported in the reserve margin. SERC expects this uncommitted capacity to continue to provide additional peaking resources for shortterm utility purchases, but the impact on the Region s 2016 summer reserve margin is uncertain. Availability of uncertain capacity cannot be assured because portions of this capacity may be designated to serve load outside of the SERC Region. Although there is no notable change in demand for the Region, some member entities report slightly decreased demand projections that they attribute to economic factors, DERs, and other energy efficiency programs. Throughout the Region, entities have various programs in place for energy efficiency and conservation: the entities incorporate the projected energy efficiency into the demand forecast, which is reflected back in their reserve margin projections, and entities can also utilize a variety of DR programs as load modifiers during high-load system conditions with implementation times that range from instantaneous to 60 minutes. Upcoming EPA policy changes related to emission standards may limit emergency stationary generator ability to operate without extensive emissions controls in response to a utility s request to reduce demand. This may have an undetermined effect for generators within the Region participating in DR programs. SERC entities coordinate transmission expansion plans in the Region annually through joint model-building efforts that include the plans of all SERC entities. The coordination of transmission expansion plans with entities outside the Region is achieved through annual participation in joint modeling efforts with the ERAG Multi-regional Modeling Working Group (MMWG). Transmission expansion plans by most SERC entities are dependent on regulatory support at the federal, state, and local levels since the regulatory entities can influence the siting, permitting, and cost recovery of new transmission lines. The regional long-term studies identified no reliability impacts due to announced retirements or significant generation outages during the assessment period. To better assess highlighted impacts in NERC s CPP Phase II Assessment from potential environmental regulations, SERC is coordinating a power-flow study utilizing the Aurora data and assumptions from the phase II assessment. In addition, several transmission upgrades are currently underway to maintain reliability within specific areas of the Region. With respect to MISO, a settlement agreement was reached between MISO, SPP, and the Joint Parties (TVA, SOCO, LG&E/KU, AECI and PowerSouth). This agreement is now in place (this superseded the ORCA on February 1, 2016) to reliably manage the magnitude of power transfers between MISO South and Midwest. The settlement agreement limits transfers between MISO-South and MISO-Midwest to 2500 MW and MISO-Midwest to MISO- South to 3000 MW in order to limit reliability impacts on neighboring systems. The increase in flow from 1,000 to 2,500/3,000 MW represents a new operating condition that has been studied and experienced under certain historical operating conditions. However, this is a significant change that will be closely monitored in operations for adverse reliability impacts. Although a settlement agreement is in place, SERC is committed to ensuring reliability of the Region and the interconnection. The Region implemented a joint loop flow study initiative with market and nonmarket entities to recreate and study loop flows within the area. The purpose of these studies is to ensure there are not potential IROL conditions that can lead to cascading, separation, or blackout conditions. SERC also plans to track and trend DERs within the Region to assess what possible impacts large penetrations of DERs may have on future BPS reliability. 139

149 Chapter 6: Regional Overview SPP The Southwest Power Pool (SPP) Planning Coordinator footprint covers 575,000 square miles and encompasses all or parts of Arkansas, Iowa, Kansas, Louisiana, Minnesota, Missouri, Montana, Nebraska, New Mexico, North Dakota, Oklahoma, South Dakota, Texas and Wyoming. The SPP Long- Term Assessment is reported based on the Planning Coordinator footprint, which touches parts of the Southwest Power Pool Regional Entity, Midwest Reliability Organization Regional Entity, and Western Electricity Coordinating Council. The SPP assessment area footprint has approximately 61,000 miles of transmission lines, 756 generating plants, and 4,811 transmission-class substations, and it serves a population of 18 million people. Summary of Methods and Assumptions Reference Margin Level SPP established target of 12.0 percent Load Forecast Method Coincident; normal weather (50/50) Peak Season Summer Planning Considerations for Wind Resources On-peak contribution of 3 percent of nameplate capacity Planning Considerations for Solar Resources On-peak contribution of 10 percent of nameplate capacity Footprint Changes The Integrated System (IS), formally part of WAPA, is reporting under the SPP assessment area this year. Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 51,936 52,819 53,235 53,409 53,779 54,336 54,703 55,119 55,581 56,048 Demand Response Net Internal Demand 51,184 51,989 52,359 52,511 52,868 53,420 53,790 54,210 54,674 55,144 Resources (MW) Anticipated 65,083 65,025 64,868 64,427 64,046 63,735 63,282 63,075 62,681 62,592 Prospective 65,004 65,925 65,768 65,497 64,775 64,665 64,212 63,901 63,101 63,011 Reserve Margins (%) Anticipated 27.16% 25.07% 23.89% 22.69% 21.14% 19.31% 17.65% 16.35% 14.65% 13.51% Prospective 27.00% 26.80% 25.61% 24.73% 22.52% 21.05% 19.37% 17.88% 15.41% 14.27% Reference Margin Level 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% 12.00% Shortfall (MW) Anticipated Prospective NERC 2016 Long-Term Reliability Assessment DRAFT October

150 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: SPP oversees the bulk electric grid and wholesale power market as one consolidated Balancing Authority area on behalf of a diverse group of utilities and transmission companies in 14 states. SPP utilized a nodal modeling technique for the probabilistic assessment. Firm imports and exports of capacity were modelled to reflect the firm transactions reported for the 2016 LTRA. Assumptions and the accompanying methodology have been thoroughly vetted through the SPP stakeholder process. No events for loss of load occurred in the Base Case and Sensitivity Case studies for the probabilistic assessment. Modeling: A Monte-Carlo-based software was used in the probabilistic assessment by randomly selecting load forecast uncertainty errors derived from historical probability of occurrence while varying the availability of thermal, hydro, and DR resources. Unit specific ramp rates, outage durations, and equivalent forced outage rates were used when varying the availability of resources in the SPP assessment area. Four thousand iterations were performed for each simulation. The generating resources modelled in the probabilistic assessment reflect the supplied data for the 2016 LTRA. Existing and Tier 1 resources were included in the probabilistic assessment along with reported confirmed retirements and projected inservice dates of new resources. Wind and solar resources were modelled at historical hourly output values. A nodal representation of transmission, load, and generation was modeled for the SPP assessment area, and transmission elements 100 kv and above were monitored to not exceed their normal rating limits. SPP flowgate and interface limitations with generation or load-related issues were considered when performing the simulations. Firm capacity transactions with firm transmission service from assessment areas external to SPP were reflected as to be continuously available during simulations, and nonfirm capacity assistance from neighboring assessment areas were not included. SPP depleted all operating reserves before shedding firm load in the probabilistic assessment. Results Trending: The 2014 Probabilistic Assessment results for SPP indicated 0.0 EUE and 0.0 hours per year LOLH for years 2016 and The 2014 Probabilistic Assessment Base Case results for 2018 were the same for the 2016 Base Case results. Also, the ProbA forecast planning reserve margin for the 2018 study year was 3 percent lower in 2014 compared to Probabilistic vs. Deterministic Reserve Margin Results: DR values reported in this report were modelled as generating resources available during daily on peak hours instead of reducing the total internal 141

151 Chapter 6: Regional Overview demand. Tier 1 wind resources with wind interconnection agreements that have not obtained firm transmission service were not included in the probabilistic assessment. Base Case Study No loss of load events were indicated for the Base Case study due to a surplus of capacity in the SPP assessment area. Reserve margins are well above 20 percent in both study years and no major impacts were observed related to resource retirements. Sensitivity Case Study No loss of load events were indicated for the Sensitivity Case study due to a surplus of capacity in the SPP assessment area. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

152 Chapter 6: Regional Overview Summary The SPP assessment area is forecasted to meet the 12 percent target reserve margin through the year The SPP assessment area s energy efficiency and conservation programs are incorporated into the reporting entities demand forecasts. There are no known impacts to the SPP assessment area s long-term reliability from energy efficiency and DR across the assessment area. The SPP assessment area forecasts the noncoincident summer peak growth at an average annual rate of one percent. The SPP assessment area studies different scenarios in short-term and long-term planning to address the impacts of renewable portfolio standards, the integration of variable resources, and the changes in resource mix. Early in 2016, the SPP assessment area saw nearly 50 percent of SPP s load being served by wind generation at certain points, setting numerous wind penetration records. SPP has been able to reliably accommodate this kind of growth so far due to its ability to anticipate it in planning efforts. SPP continues to plan transmission to meet renewable portfolio standards within the SPP assessment area. Since the previous LTRA, the SPP assessment area has not changed how on-peak capacity values for wind, solar, and hydro are calculated. The expected on-peak capacity values for variable generation are determined by guidelines established in SPP Planning Criteria section (g). 115 The SPP assessment area s 2016 Board of Directors approved SPP Transmission Expansion Plan (STEP) provides details for proposed transmission projects needed to maintain reliability while also providing economic benefit to the end users. The 2016 STEP 116 contains a comprehensive listing of all transmission projects in the SPP for the 20-year planning horizon, which consist of $6.1 billion in new transmission and upgrades. The SPP assessment area is currently not anticipating unique emerging reliability issues over the assessment time frame. However, as renewable resources continue to expand, SPP will eventually be unable to reliably utilize this generation to address internal demand needs even with additional transmission infrastructure. This will increase the need for future renewables to be delivered to other regions. SPP will continue to monitor the uncertainty of potential policy changes concerning plant retirements over the assessment period. Historically, similar to other regions, SPP has not been successful in regard to large-scale interregional transmission development. Developing the grid needed to reliably and cost-effectively accommodate expected future resource mixes will require Regions to more effectively work together to jointly plan and share costs of interregional transmission expansion. 115 SPP Planning Criteria; January SPP Transmission Expansion Plan Report; January

153 Chapter 6: Regional Overview Texas RE-ERCOT The Electric Reliability Council of Texas (ERCOT) is the Independent System Operator (ISO) for the ERCOT Interconnection and is located entirely in the state of Texas, and it operates as a single BA. ERCOT is a summer-peaking Region that covers approximately 200,000 square miles, connects 40,530 miles of transmission lines and 566 generation units, and serves 23 million customers. The Texas Reliability Entity (Texas RE) is responsible for the RE functions described in the Energy Policy Act of 2005 for the ERCOT Region. Summary of Methods and Assumptions Reference Margin Level ERCOT-established Reference Margin of percent Load Forecast Method Coincident; normal weather (50/50) Peak Season Summer Planning Considerations for Wind Resources Peak Capacity Contribution of 55 percent for Coastal units and 12 percent for Noncoastal Planning Considerations for Solar Resources ERCOT incorporates 80 percent capacity contribution Footprint Changes N/A Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 71,416 72,277 73,663 74,288 74,966 75,660 76,350 77,036 77,732 78,572 Demand Response 2,868 2,868 2,868 2,868 2,868 2,868 2,868 2,868 2,868 2,868 Net Internal Demand 68,548 69,409 70,795 71,420 72,098 72,792 73,482 74,168 74,864 75,704 Resources (MW) Anticipated 80,510 86,313 86,532 86,251 86,522 86,582 86,582 86,572 86,972 86,972 Prospective 85, , , , , , , , , ,129 Reserve Margins (%) Anticipated 17.45% 24.35% 22.23% 20.77% 20.01% 18.94% 17.83% 16.72% 16.17% 14.88% Prospective 24.07% 44.59% 46.91% 46.92% 41.86% 40.27% 38.45% 37.16% 36.42% 34.91% Reference Margin Level 13.75% 13.75% 13.75% 13.75% 13.75% 13.75% 13.75% 13.75% 13.75% 13.75% Shortfall (MW) Anticipated Prospective NERC 2016 Long-Term Reliability Assessment DRAFT October

154 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: Reserve margins for the ERCOT Region have increased since the 2014 probabilistic assessment to over 24 percent for 2018 and 20 percent in 2020, resulting in lower and insignificant levels of LOLH and EUE. Modeling: The 50/50 load forecast used for the study is based on 13 load shapes representing weather years for ERCOT applied five load forecast uncertainty multipliers (ranging from -4 percent to +4 percent) to capture load forecast uncertainty in the sequential Monte Carlo simulation model. This year s study incorporated all new hourly values for the load, wind, and hydro shapes as well as updated generator outage data. Additionally, the probabilistic model topology was simplified from six to two zones. This was done to be consistent with the ERCOT Region along with an external zone, reflecting the historical peak-period availability of capacity across five dc ties connected to SPP and Mexico. This simplification was prompted by the 2014 study results that demonstrated that internal constraints between zones had an immaterial impact on the reliability metric results. ERCOT modeled DR resources with dispatch price thresholds, call priority rankings, and availability constraints (hours-per-season and hours-per-year). Results Trending: Compared to the 2018 results for the 2014 PRA Assessment, LOLH decreased from to while EUE decreased from MWh to MWh. These reductions are due to an increase in the Anticipated Reserve Margin from 13.6 percent to 24.4 percent for the 2018 forecast year. This reserve margin increase is attributable to both a lower peak load forecast as well as an increase in anticipated resources relative to those included in the 2014 LTRA. Probabilistic vs. Deterministic Reserve Margin Results: No changes. 145

155 Chapter 6: Regional Overview Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to Planning Reserve Margins exceeding 20 percent for both forecast years. Loss of load occurred only during the summer season, with the majority in August. For example, in 2018, 78 percent of the EUE occurred in August. Relatively high values in June are driven by the 2012 weather year used to produce the load forecast. The second highest annual peak load from 2002 through 2014 occurred in June Sensitivity Case Study The results show that, as the reserve margin falls Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year) below 20 percent, the EUE remains low but begins to increase exponentially. This remains well above the target reserve margin used for the 2016 LTRA. 146

156 Chapter 6: Regional Overview Summary The Anticipated Reserve Margin is expected to remain above the Reference Margin Level (13.75 percent) for the duration of the assessment period. This is an improvement relative to the 2015 LTRA, which indicated that planning reserves would drop below the Reference Margin Level beginning in This improvement is due to a relative increase of 2,800 MW in Tier 1 capacity additions starting in 2018, which increases to over 3,800 MW for 2020 and beyond. Note that project developers typically submit interconnection requests to ERCOT no more than three to four years before the facility is expected to enter commercial operations. As a result, the Texas RE-ERCOT Region will always show a flat level of capacity additions and typically declining reserve margins starting four to five years into the LTRA forecast period. ERCOT s peak load forecast, updated in the fall of 2015, indicates system peak demand increasing at an average annual growth rate (AAGR) of approximately 1.1 percent from 2016 to Historically, summer peak demand has grown at an AAGR of 1.3 percent from 2006 to The 2016 summer peak forecast is 70,588 MW, and grows to 78,572 MW for This new peak load forecast is within about 1 percent of the one used for the 2015 LTRA, and assumes higher growth in the near- and medium-terms, but lower growth in the out years (2023 and onward). These changes in peak load growth are due to updated economic assumptions as well as an out-of-model adjustment that adds 655 MW of load to ERCOT s coastal zone to account for the expected 2019 completion of the Freeport Liquefied Natural Gas plant. ERCOT continues to rely on a variety of DR programs administered by both ERCOT and several transmission and distribution service providers (TDSPs) to support resource adequacy under emergency conditions. For Summer 2017, ERCOT estimates that it will have about 1,153 MW of load resources providing ancillary services that are contractually committed to ERCOT during summer peak hours. ERCOT also has emergency response service, a 10- and 30 minute DR and distributed generation service, designed to be deployed in the late stages of a grid emergency prior to shedding involuntary firm load. For the Summer 2016 peak hours, there are 859 MW of emergency response service contracted from 1:00 to 4:00 p.m. and 824 MW from 4:00 to 7:00 p.m. This is a six percent decrease over the same time period for Summer The most significant factor contributing to the change in participation from previous years was the U.S. Environmental Protection Agency s rule changes; these rule changes pertain to reciprocating internal combustion engines and their ability to participate in emergency DR programs. Additionally, this assessment accounts for individual TDSP contractual programs with loads that can respond to instructions to reduce total energy usage. These programs are expected to attract approximately 208 MW of additional DR capacity, and are subject to concurrent deployment with existing ERCOT DR programs, pursuant to agreements between ERCOT and the TDSPs. In aggregate, these DR programs represent 3.5 percent of the Texas RE-ERCOT Region s total internal demand forecast. Regarding new generation resources, the Texas RE-ERCOT Region saw almost 700 MW of summer-rated capacity added since the 2015 LTRA. The resource additions were dominated by wind (425 MW) and natural gas (216 MW), followed by utility-scale solar photovoltaic (72 MW). Notable new installed plants include the gas-fired Ector Country Energy Center (294 MW summer rating) and the OCI Alamo 5 solar project (95 MW nameplate, 76 MW summer rating). Additionally, there were 24 wind facilities that entered commercial service with a nameplate capacity of 3,072 MW. Notable new Tier 1 plant additions include the Colorado Bend gas combined-cycle facility (1,148 MW summer rating), Halyard Wharton Energy Center (gas peaking, 419 MW summer rating), Pinecrest G gas combined-cycle facility (785 MW summer rating), Indeck Wharton Energy Center (gas peaking, 654 MW summer rating), and Red Gate internal combustion plant (225 MW summer rating). The most significant cancelled Tier 1 project is Pondera King, an 882-MW combined-cycle plant planned for the Houston area. ERCOT does not expect that any of the six currently mothballed units will return to active status. ERCOT recently entered into a two-year Reliability Must Run (RMR) contract with an announced mothballed unit called Greens Bayou Unit 5. This 374 MW gas-steam unit is located in the Houston area, and was determined by ERCOT to be 147

157 Chapter 6: Regional Overview needed during the summer months for transmission reliability. This contract currently extends to June 30, 2018, subject to ERCOT Board of Director approval. With respect to transmission projects, the recently updated projects list includes the additions or upgrades of 3,954 miles of 138-kV and 345-kV transmission circuits, 24,159 MVA of 345/138-kV autotransformer capacity, and 3,005 MVAR of reactive capability projects that are planned in the Texas RE-ERCOT Region between 2016 and A new Houston Import Project, 130-mile 345 kv double circuit line (each circuit rated at 5000 Amps) from Limestone to Gibbons Creek to Zenith, is planned to be in service before the summer peak of The Houston area demand is met by generation located within the area and by importing power via high-voltage lines into the area from the rest of the ERCOT system. This new line will support anticipated long-term load growth in the Houston region. Power imports into the Houston area are expected to be constrained until the new import line is constructed. In July 2014, the owners of the Frontera generation plant, a 524 MW natural gas facility located on the west side of the Lower Rio Grande Valley (LRGV), announced that they were planning to switch part of the facility (170 MW) out of the ERCOT market in 2015, and the entire facility would no longer be available to ERCOT starting in the fall of In June, 2016, the ERCOT Board of Directors endorsed the reliability need for two 300 MVAr SVCs located inside the LRGV to be in service prior to Summer 2021 to meet ERCOT and NERC reliability criteria for the LRGV. ERCOT also completed under-voltage assessment and observed potential under-voltage load shedding and slow voltage recovery at 2021 summer peak load conditions in the LRGV area. Transmission upgrades were identified for this region to support the voltage recovery and meet the NERC and ERCOT planning criteria by year The upgrades include two 300 MVAR Static VAR Compensators (SVCs) in the LGRV area. In June 2016, the ERCOT Board of Directors endorsed the reliability need for these two 300 MVAr SVCs. The Texas Panhandle region is currently experiencing significantly more interest from wind generation developers than what was initially planned for the area. The ERCOT Panhandle grid is remote from synchronous generators and requires long distance power transfer to the load centers in the Texas RE-ERCOT Region. All wind generation projects in the Panhandle are expected to be equipped with advanced power electronic devices that will further weaken the system due to limited short-circuit current contributions. Stability challenges and weak system strength are expected to be significant constraints for Panhandle export. The ERCOT Transmission Planning Department has been performing ongoing analysis to assess reliability when incorporating all wind generation in the Panhandle that satisfy the requirements of ERCOT s transmission planning process. The stability and system strength are evaluated to ensure that reliable operations can be maintained through proper implementation of Panhandle export limits. In terms of long-term resource adequacy and reliability risks, the retirement of multiple coal-fired generating units during the assessment period due to federal environmental regulations and market economics remains the greatest known risk. A number of coal units in the Texas RE-ERCOT Region are at risk for retirement due to requirements to upgrade existing flue gas desulfurization (FGD) equipment or install new equipment under the EPA s Texas Federal Implementation Plan (FIP) for regional haze. Under the Texas FIP, 12 coal-fired units in ERCOT (totaling about 8,500 MW) will require FGD investments either by February 2019 (for units requiring upgrades) or February 2021 (for units requiring new equipment). In June 2016, the U.S. Court of Appeals for the 5th Circuit placed a judicial stay on implementation of the EPA's rule, thereby likely postponing the need for the unit owners to invest in new scrubbers or scrubber upgrades. As a result, there is large uncertainty regarding if and when coal units are retired. Nevertheless, ERCOT s transmission reliability study, conducted in the fall of 2015 to analyze the impacts of multiple coal unit retirement scenarios, indicated that local or regional transmission impacts (transmission and transformer overloads) would be expected for all scenarios. To address transmission issues caused by specific retiring units, ERCOT and its stakeholders would pursue necessary transmission infrastructure upgrades or other alternatives (such as installation of voltage control devices or interruptible load procurement) through ERCOT s transmission planning and project review process. 148

158 Chapter 6: Regional Overview WECC The Western Electricity Coordinating Council (WECC) is responsible for coordinating and promoting BPS reliability in the Western Interconnection. WECC s 329 members, which include 38 BAs, represent a wide spectrum of organizations with an interest in the BPS. Serving an area of nearly 1.8 million square miles and approximately 82.2 million people, it is geographically the largest and most diverse of the NERC regional reliability organizations. WECC s service territory extends from Canada to Mexico. It includes the provinces of Alberta and British Columbia in Canada, the northern portion of Baja California in Mexico, and all or portions of the 14 western states in between. The WECC assessment area is divided into five subregions: Rocky Mountain Reserve Group (RMRG), Southwest Reserve Sharing Group (SRSG), California/Mexico (CA/MX), and the Northwest Power Pool (NWPP), which is further divided into the BC, AB, and NWPP-US areas. These subregional divisions are used for this study as they are structured around Reserve Sharing groups that have similar annual demand patterns and similar operating practices. Summary of Methods and Assumptions Reference Margin Level Determined by WECC s building block method for each subregion. Load Forecast Method Coincident for each subregion; normal weather (50/50) Peak Season Summer: CA/MX, RMRG, SRSG, and NWPP-US Winter: AB and BC Planning Considerations for Wind Resources Modeling, primarily based on up to five years of historic data Planning Considerations for Solar Resources Modeling, primarily based on up to five years of historic data Footprint Changes N/A WECC-AB Assessment Area Footprint WECC-Total Footprint WECC-BC Assessment Area Footprint NERC 2016 Long-Term Reliability Assessment DRAFT October

159 Chapter 6: Regional Overview WECC-CA/MX Assessment Area Footprint WECC-NWPP-US Assessment Area Footprint WECC-RMRG Assessment Area Footprint WECC-SRSG Assessment Area Footprint WECC-AB Peak Season Demand, Resources, Reserve Margins, and Shortfall 117 Demand (MW) Total Internal Demand ,942 13,198 13,460 13,705 13,910 14,114 14,304 Demand Response Net Internal Demand ,942 13,198 13,460 13,705 13,910 14,114 14,304 Resources (MW) Anticipated ,287 16,439 16,366 16,235 16,282 16,287 16,424 Prospective ,648 19,902 19,788 19,565 19,643 19,648 19,878 Reserve Margins (%) Anticipated 33.56% 32.80% 29.62% 25.84% 24.56% 21.59% 18.46% 17.05% 15.39% 14.82% Prospective 43.44% 54.86% 53.04% 51.81% 50.80% 47.01% 42.76% 41.22% 39.21% 38.97% Reference Margin Level 11.03% 11.03% 11.03% 11.03% 11.03% 11.03% 11.03% 11.03% 11.03% 11.03% Shortfall (MW) Anticipated Prospective Per WECC's request, data is not presented publically for Alberta and British Columbia subregions. 150

160 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-AB is a winter-peaking system that covers the province of Alberta Canada. For the probabilistic assessment, WECC utilized the Multi-Area Variable Resource Integration Convolution (MAVRIC) model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits, consistent with the LTRA, based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, and geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s Generation Availability Data System (GADS). Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as Demand Distributions. Results Trending: The previous assessment showed slightly greater than zero LOLH and EUE; however, the results were insignificant and below the reporting threshold. This year s assessment reflected a decrease in expected demand as compared to previous assessments resulting in LOLH and EUE of near zero. Probabilistic vs. Deterministic Reserve Margin Results: The probabilistic assessment reserve margins are based on the transfer capability between regions whereas the LTRA deterministic reserve margins are based on transfers needed to maintain reference margins. Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to planning reserve margins exceeding 33 percent and 29 percent for 2018 and 2020 respectively. Sensitivity Case Study The EUE and LOLH remain nil for the Sensitivity Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

161 Chapter 6: Regional Overview WECC-BC Peak Season Demand, Resources, Reserve Margins, and Shortfall 118 Demand (MW) Total Internal Demand ,140 12,242 12,401 12,524 12,690 12,853 13,040 Demand Response Net Internal Demand ,140 12,242 12,401 12,524 12,690 12,853 13,040 Resources (MW) Anticipated ,642 13,757 13,935 14,073 14,260 14,344 14,316 Prospective ,642 13,757 13,935 14,073 15,250 15,334 15,306 Reserve Margins (%) Anticipated 12.39% 12.39% 12.38% 12.37% 12.38% 12.37% 12.37% 12.37% 11.60% 9.79% Prospective 12.39% 12.39% 12.38% 12.37% 12.38% 12.37% 12.37% 20.17% 19.31% 17.38% Reference Margin Level 12.10% 12.10% 12.10% 12.10% 12.10% 12.10% 12.10% 12.10% 12.10% 12.10% Shortfall (MW) Anticipated Prospective Per WECC's request, data is not presented publically for Alberta and British Columbia subregions. 152

162 Chapter 6: Regional Overview Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-BC is a winter-peaking system that covers the province of British Columbia Canada. For the probabilistic assessment, WECC utilized the MAVRIC model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits consistent with the LTRA based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, and geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s GADS. Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as demand distributions. Results Trending: The previous assessment showed slightly greater than zero LOLH and EUE however the results were insignificant and below the reporting threshold. This year s assessment reflected a decrease in expected demand as compared to previous assessments resulting in LOLH and EUE of zero. Probabilistic vs. Deterministic Reserve Margin Results: The probabilistic assessment reserve margins are based on the transfer capability between regions where as the LTRA deterministic reserve margins are based on transfers needed to maintain reference margins. Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to planning reserve margins exceeding 23% and 20% for 2018 and 2020 respectively. Sensitivity Case Study The EUE and LOLH remain nil for the Sensitivity Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

163 Chapter 6: Regional Overview WECC-CA/MX Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 54,774 54,554 54,335 54,221 54,162 54,287 54,301 54,260 54,129 54,005 Demand Response 1,747 1,706 1,707 1,707 1,707 1,707 1,707 1,707 1,707 1,707 Net Internal Demand 53,027 52,848 52,628 52,514 52,455 52,580 52,594 52,553 52,422 52,298 Resources (MW) Anticipated 63,765 64,428 64,261 63,711 63,626 63,441 63,364 63,068 62,796 61,639 Prospective 63,859 64,522 64,427 63,912 63,827 63,642 63,565 62,139 60,736 59,898 Reserve Margins (%) Anticipated 20.25% 21.91% 22.10% 21.32% 21.30% 20.66% 20.48% 20.01% 19.79% 17.86% Prospective 20.43% 22.09% 22.42% 21.70% 21.68% 21.04% 20.86% 18.24% 15.86% 14.53% Reference Margin Level 16.16% 16.16% 16.16% 16.16% 16.16% 16.16% 16.16% 16.16% 16.16% 16.16% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-CAMX is a summer-peaking system that covers the state of California and portions of Baja Mexico. For the probabilistic assessment, WECC utilized the MAVRIC model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits consistent with the LTRA based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, and geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s GADS. Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as Demand Distributions. Results Trending: The previous assessment showed slightly greater than zero LOLH and EUE however the results were insignificant and below the reporting threshold. This year s assessment reflected a decrease in expected demand as compared to previous assessments resulting in LOLH and EUE of zero. 154

164 Chapter 6: Regional Overview Probabilistic vs. Deterministic Reserve Margin Results: The probabilistic assessment reserve margins are based on the transfer capability between regions where as the LTRA deterministic reserve margins are based on transfers needed to maintain reference margins. Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to planning reserve margins exceeding 32 percent and 36 percent for 2018 and 2020 respectively. Sensitivity Case Study The EUE and LOLH remain nil for the Sensitivity Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

165 Chapter 6: Regional Overview WECC-NWPP-US Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 50,013 50,438 50,851 51,319 51,693 52,119 52,479 52,869 53,439 53,294 Demand Response 1,219 1,186 1,196 1,195 1,195 1,193 1,213 1,189 1,187 1,193 Net Internal Demand 48,794 49,252 49,655 50,124 50,498 50,926 51,266 51,680 52,252 52,101 Resources (MW) Anticipated 62,374 64,519 65,385 65,313 64,879 64,453 65,266 65,446 65,505 65,079 Prospective 62,568 64,913 65,927 65,811 65,288 64,819 65,792 65,987 66,003 65,562 Reserve Margins (%) Anticipated 27.83% 31.00% 31.68% 30.30% 28.48% 26.56% 27.31% 26.64% 25.36% 24.91% Prospective 28.23% 31.80% 32.77% 31.30% 29.29% 27.28% 28.34% 27.68% 26.32% 25.84% Reference Margin Level 16.32% 16.32% 16.32% 16.32% 16.32% 16.32% 16.32% 16.32% 16.32% 16.32% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-NWUS is a summer-peaking system that covers a triangle of states from Washington to Montana and down through Nevada. For the probabilistic assessment, WECC utilized the MAVRIC model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits consistent with the LTRA based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, and geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s GADS. Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as Demand Distributions. Results Trending: The previous assessment showed slightly greater than zero LOLH and EUE however the results were insignificant and below the reporting threshold. This year s assessment reflected a decrease in expected demand as compared to previous assessments resulting in LOLH and EUE of zero. 156

166 Chapter 6: Regional Overview Probabilistic vs. Deterministic Reserve Margin Results: The probabilistic assessment reserve margins are based on the transfer capability between regions where as the LTRA deterministic reserve margins are based on transfers needed to maintain reference margins. Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to planning reserve margins exceeding 41 percent and 37 percent for 2018 and 2020 respectively Sensitivity Case Study The EUE and LOLH remain nil for the Sensitivity Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

167 Chapter 6: Regional Overview WECC-RMRG Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 12,392 12,530 12,759 12,947 13,194 13,375 13,552 13,739 13,910 14,094 Demand Response Net Internal Demand 11,847 11,968 12,178 12,344 12,585 12,760 12,932 13,113 13,279 13,459 Resources (MW) Anticipated 15,364 15,018 15,026 15,022 15,230 15,338 15,539 15,711 15,879 16,088 Prospective 15,364 14,975 14,945 14,951 15,159 15,270 15,472 15,685 15,879 16,088 Reserve Margins (%) Anticipated 29.68% 25.49% 23.39% 21.69% 21.02% 20.20% 20.16% 19.81% 19.58% 19.53% Prospective 29.68% 25.13% 22.72% 21.12% 20.45% 19.67% 19.64% 19.61% 19.58% 19.53% Reference Margin Level 14.14% 14.14% 14.14% 14.14% 14.14% 14.14% 14.14% 14.14% 14.14% 14.14% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-RMRG is a summer peaking system that covers the states of Wyoming and Colorado. For the probabilistic assessment, WECC utilized the MAVRIC model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits consistent with the LTRA based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, and geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s GADS. Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as Demand Distributions. Results Trending: The previous assessment showed slightly greater than zero LOLH and EUE however the results were insignificant and below the reporting threshold. This year s assessment reflected a decrease in expected demand as compared to previous assessments resulting in LOLH and EUE of zero. 158

168 Chapter 6: Regional Overview Probabilistic vs. Deterministic Reserve Margin Results: The probabilistic assessment reserve margins are based on the transfer capability between regions where as the LTRA deterministic reserve margins are based on transfers needed to maintain reference margins. Base Case Study For the Base Case study, EUE and LOLH values were insignificant due to planning reserve margins exceeding 39 percent and 34 percent for 2018 and Sensitivity Case Study The EUE and LOLH remain nil for the Sensitivity Case. Summary of Results Reserve Margin (RM) % Base Case Sensitivity Case Anticipated Prospective Reference ProbA Forecast Planning ProbA Forecast Operable Annual Probabilistic Indices Base Case Sensitivity Case EUE (MWh) EUE (ppm) LOLH (hours/year)

169 Chapter 6: Regional Overview WECC-SRSG Peak Season Demand, Resources, Reserve Margins, and Shortfall Demand (MW) Total Internal Demand 23,207 23,534 24,023 24,479 24,978 25,319 25,766 26,211 26,871 27,424 Demand Response Net Internal Demand 22,787 23,175 23,660 24,124 24,623 24,964 25,411 25,856 26,516 27,069 Resources (MW) Anticipated 29,094 29,339 29,289 29,231 29,182 29,771 29,995 30,422 31,136 31,763 Prospective 29,095 28,504 28,454 29,395 29,346 29,936 30,160 30,586 31,300 31,927 Reserve Margins (%) Anticipated 27.68% 26.60% 23.79% 21.17% 18.52% 19.26% 18.04% 17.66% 17.42% 17.34% Prospective 27.68% 22.99% 20.26% 21.85% 19.18% 19.92% 18.69% 18.29% 18.04% 17.95% Reference Margin Level 15.82% 15.82% 15.82% 15.82% 15.82% 15.82% 15.82% 15.82% 15.82% 15.82% Shortfall (MW) Anticipated Prospective Peak Season Reserve Margins On-Peak Tier 1 Capacity Additions Probabilistic Assessment Overview General Overview: WECC-SRSG is a summer peaking system that covers the states of New Mexico and Arizona and a portion of California. For the probabilistic assessment, WECC utilized the MAVRIC model. MAVRIC utilizes the convolution method of probabilistic studies. It is designed to study all possible subsets of probability distributions associated with demand, generation, and transmission to determine the extent of possible loss of load in each of the systems. Transmission Modeling: Each Balancing Authority was modeled with import and export limits consistent with the LTRA based on expected power flow transfers. Unit Modeling: For base-load resources (nuclear, thermal, geothermal), WECC utilized unit specific outage, planning, and maintenance outage rates within the probabilistic assessment that were based on 10 years of data from NERC s GADS. Seven years of historical hourly generation was utilized to develop availability probability distributions for renewable resources (hydro, wind, and solar) as well as Demand Distributions. 160